L Methionine PDF

Production of L-methionine with
Corynebacterium glutamicum

Von der Fakultt fr Lebenswissenschaften
der Technischen Universitt Carolo Wilhelmina
zu Braunschweig
zur Erlangung des Grades eines
Doktors der Naturwissenschaften
(Dr. rer. nat)
genehmigte
Dissertation

von Hajo Kampe Reershemius
aus Norden (Ostfriesland)

1. Referent: Prof. Dr. Siegmund Lang
2. Referent: Prof. Dr. Klaus Dieter Vorlop
eingereicht am: 25.06.2008
mndliche Prfung (Disputation) am: 14.11.2008
Druckjahr 2008

Danksagung
___________________________________________________________________________

Danksagung

Herrn Prof. Dr. Siegmund Lang danke ich fr die interessante Aufgabenstellung und die Mglichkeit
der Durchfhrung meiner Doktorarbeit am Institut fr Biochemie und Biotechnologie sowie die ber-
nahme des Hauptreferates. Seine Untersttzung und sein Ratschlag waren von groem Wert, vor allem
in schwierigeren Phasen whrend der Durchfhrung meiner Arbeit.

Herrn Prof. Dr. K. D. Vorlop gilt mein Dank fr die bernahme des Korreferats und die Untersttzung
von Seiten seines Institutes (Heinrich von Thnen Institut, Institut fr Agrartechnologie).

Mein Dank gilt der japanischen Firma Kyowa Hakko Kogyo Ltd., Tokio, fr die Bereitstellung von
mikrobiellem Material fr meine wissenschaftliche Arbeit.

Bei Herrn Wolfgang Grassl mchte ich mich fr die Untersttzung bei der Durchfhrung der Bioreak-
torexperimente bedanken.

Bei Herrn Hasan Cicek mchte ich mich fr die Hilfestellung bei der Einfhrung in die Gaschroma-
tographie Metechnik bedanken.

Herrn Dr. Till Beuerle danke ich fr die Untersttzung bei der Durchfhrung und Interpretation der
GC MS Untersuchungen.

Frau Ileana Jurchescu mchte ich ganz besonders fr Ihre tatkrftige Untersttzung bei den wissen-
schaftlichen Untersuchungen danken.

Bedanken mchte ich mich bei meinen Kollegen Andrea Holtkamp, Ariane Schwoerer, Julika Wren-
ger, Olof Palme, Rolf Heckmann, Andrea Walzog, Linda Kilian und Malte Timm fr die angenehme
Arbeitsatmosphre.

Bei meiner Schwester Gertrud Reershemius mchte ich mich fr das Korrekturlesen der englischen
Sprache bedanken.

Bei meinen Eltern, Gerda und Kampe Reershemius, mchte ich mich fr die vielseitige Untersttzung
bedanken, die diese Arbeit erst ermglicht hat.

Abstract
___________________________________________________________________________
Abstract
Targets of this project were the screening for microbial overproducers of L-methionine by random
mutation (UV-radiation plus following selection) and optimization of the production in shake flask
and bioreactor scale by varying of biochemical (nutrient concentration, pH) and general cultivation
parameters (temperature, stirring speed, in bioreactor: oxygen partial pressure and fed batch).

In this context one major focus of the research was the development of a fast screening and selection
system for microbial overproducers of L-methionine. For the necessary experiments the strain Cory-
nebacterium glutamicum DSM20300 was used. A big problem was the selection system after radiation
with ultraviolet light (UV) to sort out potential candidates for the analysis. Although a successful inhi-
bition of non-methionine producing microorganisms using methionine analogues like ethionine had
been described in former experiments (Mondal et al. 1996), it was not possible to reproduce these
results. Due to this problem in the selection method the measurement of a huge amount of colonies
obtained after the UV-radiation could not be avoided. In spite of all efforts undertaken to accelerate
the analysis methods, the chances to find overproducers in screening experiments were noticeable
diminished. The attempt to find a microbial overproducer in these screening experiments has not been
successful.

A second challenge was the searching for microorganisms already containing the property to overpro-
duce L-methionine. One strain, Corynebacterium glutamicum ATCC21608, is commercially available
from a public collection of microorganisms in the USA. It could be shown in several experiments, that
this bacterium is able to produce L-alanine in small amounts (200 mg/l) but L-methionine could not be
produced in significant quantities.

The strain Corynebacterium glutamicum KY10574 has been obtained from the company Kyowa
Hakko Kogyo, Japan, a well-known producer of amino acids with microorganisms. A successful ac-
cumulation of L-methionine was observed. In first experiments the concentration of L-methionine in
the supernatant of the cultivations was about 100 150 mg/l, in further experiments in the shake flask
(several scales) and the bioreactor (3 3.5 litres) L-methionine concentrations of 1.4 1.5 g/l could be
achieved. For both bioreactor and shake flask scale significant parameters for this optimization were
the creation of an adequate minimal medium and the oxygen supply in combination with the stirring
speed. In bioreactor scale the best process was the cultivation with a limitation factor (sugar, glucose)
in order to be able to control cell growth velocity and guarantee aerobic conditions through efficient
oxygen supply. L-methionine was only produced in higher amounts under aerobic conditions in the
cultivation broth.

Abstract
___________________________________________________________________________
Abstract
Ziel dieses Projektes war das Screening nach mikrobiellen berproduzenten der Aminosure L-
Methionin durch zufallsbasierte Mutagenese (UV-Bestrahlung und nachfolgende Selektion). Darauf-
hin sollte die Produktion im Schttelkolben- und Bioreaktormastab durch Variation der biochemi-
schen (Nhrstoffkonzentrationen, pH) und generellen Kultivierungsparameter (Temperatur, Schttel-
und Drehzahlraten, im Bioreaktor: Sauerstoffpartialdruck und Zuftterungsstrategien) optimiert wer-
den.

Zunchst war die Entwicklung eines schnellen und effizienten Screenings- und Selektionssystems fr
mikrobielle berproduzenten von L-Methionin das Hauptziel des Projektes. Dafr wurde der Stamm
Corynebacterium glutamicum DSM20300 benutzt. Als grosses Problem stellte sich die Etablierung
eines effizienten Selektionssystem zum Aussortieren potentieller Kandidaten fr die Aminosureana-
lyse heraus. Obwohl eine erfolgreiche Inhibierung mikrobiellen Wachstums mit Methioninanaloga wie
Ethionin in der Literatur beschrieben worden ist (Mondal et al. 1996), war es nicht mglich diese Er-
gebnisse zu reproduzieren. Aufgrund dieser Probleme mit der Selektionsmethode konnte das Vermes-
sen einer grossen Menge von Mutantenkolonien, die nach UV-Bestrahlung erhalten worden sind, nicht
vermieden werden.

Trotz aller Versuche die Analytik zu beschleunigen, waren die Chancen einen berproduzenten in den
Screeningexperimenten zu finden, dadurch sehr verringert. Es war im weiteren Verlauf der Experi-
mente nicht mglich einen L-Methionin berproduzenten herzustellen.

Ein weiterer Punkt war die Suche nach Mikroorganismen, die bereits die Fhigkeit zur Produktion von
Methionin besitzen. Der Stamm Corynebacterium glutamicum ATCC21608 ist kommerziell an einer
amerikanischen Stammsammlung erhltlich. In verschiedenen Versuchen konnte gezeigt werden, dass
dieser Stamm die Fhigkeit zur L-Alaninproduktion in Mengen um die 200 mg/l besitzt. L-Methionin
wurde von diesem Stamm nicht berproduziert.

Der Stamm Corynebacterium glutamicum KY10574 wurde von der Firma Kyowa Hakko Kogyo, Ja-
pan, Tokio zur Verfgung gestellt. Hier konnte eine L-Methioninproduktion beobachtet werden. In
den ersten Untersuchungen konnten L-Methionin Konzentrationen zwischen 100 150 mg/l im ber-
stand gemessen werden, in weiteren Experimenten im Schttelkolben- und Bioreaktormastab wurden
Konzentrationen zwischen 1,4 und 1,5 g/l erzielt. Wichtige Parameter fr diese Optimierung waren die
Herstellung eines adquaten Minimalmediums, die Sauerstoffversorgung und die Zuftterungsstrate-
gie (Glucose) im Bioreaktor. L-Methionin wurde nur bei ausreichendem Sauerstofftransfer in das Kul-
tivierungsmedium in hheren Mengen produziert.

Table of contents
___________________________________________________________________________
1 Introduction ...................................................................................... 1
2 Theoretical background................................................................... 3
2.1 Amino Acid Production ..................................................................... 3
2.2 L-methionine....................................................................................... 4
2.2.1 DL-methionine production by chemical synthesis ................................. 6
2.2.2 L-methionine production by enzymatic synthesis.................................. 7
2.3 Corynebacterium glutamicum........................................................... 8
2.3.1 General information............................................................................... 8
2.3.2 L-glutamic acid production with Corynebacterium glutamicum............ 10
2.3.3 L-lysine production with Corynebacterium glutamicum....................... 11
2.3.4 L-methionine production with Corynebacterium glutamicum............... 13
2.3.5 Amino acid excretion mechanisms in Corynebacterium glutamicum... 19
2.4 Biochemical amino acid degradation............................................. 20
2.5 Random mutagenesis and following selection ............................. 22
2.6 Amino Acid Analysis........................................................................ 24
3 Materials and methods .................................................................. 26
3.1 Microbial strains............................................................................... 26
3.2 Media and biotransformation experiments with Arthrobacter
aurescens DSM7330................................................................................... 28
3.3 Corynebacterium glutamicum media............................................. 29
3.3.1 Conservation media............................................................................ 29
3.3.2 Media for Corynebacterium glutamicum DSM20300........................... 30
3.3.3 Media for Corynebacterium glutamicum ATCC21608......................... 32
3.3.4 Media for Corynebacterium glutamicum KY10574.............................. 32
3.4 Analysis techniques......................................................................... 35
3.4.1 Bio dry mass and optical density......................................................... 35
3.4.1.1 Connection between the BDM and OD
546
values ................................................... 35
3.4.2 pH-measurement ................................................................................ 36
3.4.3 Glucose measurement........................................................................ 36
3.4.4 Thin layer chromatography (TLC) ....................................................... 39
3.4.5 High pressure / performance liquid chromatography (HPLC).............. 41
3.4.6 Gas chromatography (GC).................................................................. 42
3.4.6.1 Gas chromatography - mass spectrometry (GC-MS) ............................................. 45
3.5 Bioreactor experiments ................................................................... 45
3.5.1 Exhaust gas analysis .......................................................................... 48
3.6 Screening and selection system..................................................... 49
3.7 Cell disruption of C. glutamicum strains ....................................... 51
3.7.1 Bradford assay.................................................................................... 52
3.8 Protein hydrolysis of C. glutamicum KY10574.............................. 52
3.9 Chemicals and equipment ............................................................... 52
4 Results and discussion ................................................................. 53
4.1 Arthrobacter aurescens DSM7330: L-methionine production
with biotransformation .............................................................................. 53
4.2 Corynebacterium glutamicum DSM20300 ..................................... 55
- I -
Table of contents
___________________________________________________________________________
4.2.1 L-glutamic acid production with C. glutamicum DSM20300................ 55
4.2.2 L-methionine production capability of C. glutamicum DSM20300....... 56
4.2.3 Screening for L-methionine overproducers with C. glutamicum
DSM20300 with random mutagenesis and following selection.......................... 58
4.3 Corynebacterium glutamicum ATCC21608 ................................... 61
4.4 Corynebacterium glutamicum KY10574 ........................................ 67
4.4.1 First cultivation in the shake flask scale.............................................. 67
4.4.1.1 Use of GC-MS amino acid identification for KH1 / KH2 media cultivation.............. 71
4.4.2 Shake flask cultivation of Corynebacterium glutamicum KY10574
in F1 minimal medium....................................................................................... 77
4.4.3 Shake flask cultivations in F1 minimal medium with different
glucose levels at rising shaking rates................................................................ 81
4.4.3.1 25 ml cultivation (250 ml shake flask) with 20 g/l glucose...................................... 82
4.4.3.5 25 ml cultivation (250 ml shake flask) with 100 g/l glucose.................................... 88
4.4.3.6 100 ml cultivation (1 L shake flask) with 60 g/l glucose.......................................... 91
4.4.3.7 Comparison of the L-methionine production with C.glutamicum KY10574 with
different glucose amounts and shake flask scales................................................................... 92
4.4.4 Shake flask cultivation of C. glutamicum KY10574 in F1 medium
with different shaking rates................................................................................ 95
4.4.5 Shake flask cultivation in F1 minimal medium using precultures......... 96
4.4.6 Shake flask cultivation of Corynebacterium glutamicum KY10574
using metabolic precursors of L-methionine...................................................... 98
4.4.7 Cultivation of Corynebacterium glutamicum KY10574 in the
bioreactor scale (3.5 L).....................................................................................100
4.4.7.1 Bioreactor cultivation with 10 g/l glucose at t =0 and a continuous feeding
strategy for additional glucose with high pumping rates ........................................................ 100
strategy for additional glucose with low pumping rates.......................................................... 105
strategy for additional glucose with low pumping rates.......................................................... 109
strategy for additional glucose with low / mid pumping rates and an effective pO
2
control
system .............................................................................................................................. 114
4.4.7.5 Summary about the bioreactor cultivations with Corynebacterium glutamicum
KY10574 .............................................................................................................................. 119
4.4.8 Investigation of total L-methionine concentrations (extra- and
intracellular) of the strain C. glutamicum KY10574 compared to the strain
C. glutamicum DSM20300................................................................................122
4.4.8.1 Protein hydrolysis of C. glutamicum KY10574 bio dry mass ................................ 123
5 Summary / Outlook ...................................................................... 126
6 References.................................................................................... 128
7 Abbreviations ............................................................................... 134
8 Attachment ................................................................................... 135
9. Curriculum vitae / Lebenslauf ....................................................... 140
- II -
Introduction
___________________________________________________________________________
1 Introduction
Apart from l-lysine and l-threonine, L-methionine belongs to the essentiell amino acids which
human and animal metabolisms are not capable to produce. Most natural feeds as wheat or
maize protein, soya bean and fish meal are deficient in methionine, lysine and threonine.
Therefore these amino acids are the most important nutrient additives in animal feeding. For
pig and poultry breeding specific feeding plans exist with L-methionine as essentiell and sul-
phur containing amino acid. The impact of L-methionine on animal nutrition and the conse-
quences of its absence as nutritive feed additive have been investigated very well. It has been
observed for poultry that the stability of egg shells decreases just as the milk production in
cows does (Noftsger et al. 2003; Keshavarz et al. 2003).

In 1995 the demand for methionine amounted to 300,000 tons per year (Leuchtenberger et al.
1996). The general and cheapest process to obtain L-methionine is the chemical synthesis
using acroleine, methyl mercaptan and hydrocyanic acid (Leuchtenberger 1996; Pack 2004).
The product is a racemic mixture of D- and L-methionine. Another method to obtain L-
methionine is the extraction from protein hydrolysates (Kircher et al. 1998). In 1998 500,000
tons D/L-methionine per year were produced (Toride 2002), the market for amino acids in
general and for L-methionine in special increases constantly.

Normally only the L-form can be utilized by human and animal metabolisms, but for me-
thionine enzymes are available in human and animal bodies which make it possible to convert
the D-form of the chemically synthesized racemic DL-methionine mixture. The organism
converts the D-form enzymatically into the nutritive L-form via an amino acid oxidase system
by oxidative desamination and subsequent transamination reactions (Leuchtenberger et al.
2005; Hasegawa et al. 2005). This is an enormous potential for the manufacturer to reduce
costs and the reason why there is no other process to produce methionine and replace the
chemical production up to the present.

However, a new EU ordinance from August 2005 prohibits the use of synthesized methionine
for animal feed addition in ecological farming. Due to this law new alternative methods to
obtain methionine from ecological resources will need to be developed. All current work
about de novo synthesis of L-methionine in the European Union with renewable primary
products like sugar and starch are based on genetic engineering (directed mutations) of the

- 1 -
Introduction
___________________________________________________________________________
metabolic pathway and the modification of the enzyme activities and the intermediates / pre-
cursors which occur in the metabolic pathway of L-methionine (e.g. Mampel et al. 2005).

A model organism for the overproduction of amino acids in microorganisms is Corynebacte-
rium glutamicum, because of its simplified metabolic pathways for the production of amino
acids (Lee et al. 2003; Gomes et al. 2005). For this reason it is easier to develop a new strain
with changed metabolic fluxes in order to produce L-methionine. The production of L-lysine,
another important animal food additive, is performed also with overproducing strains of
Corynebacterium glutamicum (Eggeling et al. 1999). For L-lysine, it is the state of the art
technique.

Most publications about L-methionine production mention about the screening system which
leads to microbial overproducer mutants. The problem is, that the production of amino acids
in microorganisms is subject to a feedback inhibition, which prevents overproduction
(Wartenberg 1989). For this reason the first target in research is to find a way to deactivate
this molecular mechanism. The first and traditional method is the random mutation with fol-
lowing selection (Gerhardt et al. 1981). The treatment of cells with UV-radiation or chemical
mutation agents (random mutation) and the selection with methionine analogues / antimetabo-
lites is widely described in the literature (Kase et al. 1974; Mondal et al. 1996; Kumar et al.
2003).

Over the last 20 years, genetic engineering with directed mutation methods has developed into
another approach with increasing potential. The knowledge of pathways and mechanisms
from gene to product allows to work specifically through the activating or inactivating of
genes and enzyme activities. Due to gene technology it is possible to change the metabolic
flux in microorganisms step by step and canalise it in the designated direction to reach sig-
nificant amounts of a special metabolite.

Today it is common to make attempts to get microbial overproducers with both random and
directed mutation methods. The experiences in the production of other amino acids like L-
lysine and L-glutamic acid (e.g. Eggeling et al. 1999) will be of great value for the research
on L-methionine production in the European Union (EU) in the next years in order to develop
microbial overproducers and economical biotechnical processes for L-methionine production.

- 2 -
Theoretical background
___________________________________________________________________________
2 Theoretical background
2.1 Amino Acid Production

The production of amino acids is a big industrial factor in both the chemical and biotechno-
logical industries. There has been always hard competition between these two fields to pro-
duce amino acids in a cheap and energy reducing mode.

Amino acids have many special properties which make them very valuable, as for example
their contribution to nutrition, the taste, the chemical features and their importance in physio-
logical activities. The proteinogenic amino acids are the building blocks of proteins, they are
important intermediates on the pathway from the genetic to the protein level.

The varied use of amino acids are as supplements to human and animal food, medical infu-
sions, cosmetics and intermediates in the chemical industry. According to data from 1995 the
whole market is estimated at 3 billion US $ in 1995. divided in 38% for food, 54% for feed
and 8% for other applications (Leuchtenberger et al. 1996). Figure 2.1 shows an overview of
all methods to gain amino acids in industry.

C
R
NH
2
COOH
H
chemical synthesis protein hydrolysis (extraction)
enzymatic synthesis
fermentation, cultivation
(microbial overproducers)

Figure 2.1: Principle possibilities to produce L-amino acids or D/L-amino acids in the case of chemi-
cal synthesis (modified according to Leuchtenberger et al. 1988).

- 3 -
___________________________________________________________________________
It is possible to synthesize all amino acids in the traditional chemical way but for many of
them it would be much more profitable to produce with different methods. The advantage of
the enzymatic synthesis and the direct fermentation is the modern enantioselective production
of either the L- or D-enantiomeric form. There are examples for each of the production possi-
bilities mentioned in Figure 2.1. Glycine is the only nonchiral amino acid, therefore the
chemical process is without competition because there is no racemic product mixture to pu-
rify. For L-methionine, the chemical synthesis in combination with the enzymatic resolution
of the racemic mixture is the most important form of production.

L-asparagine, L-arginine, L-histidine and L-cysteine for example are produced by extraction
from protein hydrolysates; L-tryptophan and L-aspartic acid are obtained using enzymes or
immobilized cells.

The barrier for multi enzyme systems is reached when the effectiveness of the microbial cell
as enzyme membrane reactor is much higher in spite of side reactions and by-products. On
this account the direct fermentation is the preferable process in commercial aspects for L-
lysine and L-glutamic acid (Kole et al. 1986; Kinoshita et al. 1961; Kiefer et al. 2004).

A major problem is the strong regulated biosynthesis in wild type microorganisms. The pro-
duced amino acid itself restricts the formation of necessary enzymes (feedback repression)
and / or reduces the activity of key enzymes for the metabolic building pathway (feedback
inhibition) (Leuchtenberger et al. 1988). In a suitable strain the control mechanisms have to
be deactivated. In addition, side reactions and the degradation of end and intermediate prod-
ucts have to be blocked. The export and discharge of the product to the extracellular environ-
ment needs to be considered, too (Krmer 1993; Trtschel et al. 2005).
2.2 L-methionine

Methionine (CAS registry number 63-68-3) occurs in two enantiomeric chiral forms. L-
methionine is the proteinogenic one. It has nonpolar properties and is, apart from cysteine, the
unique amino acid with a sulphur component. In normal environment conditions it is a solid
white powder. Other physicochemical properties and the molecular structure are shown in the
Table 2.1 and Figure 2.2. According to the metabolic pathway L-methionine belongs together
with L-lysine and L-threonine to the aspartic group of amino acids. These 3 proteinogenic

- 4 -
___________________________________________________________________________
acids, which are all of great importance for human and animal food, become synthesized from
aspartic acid, the first corporate reaction is catalyzed by aspartate kinase, an enzyme only oc-
curring in plants and microorganisms (Voet et al. 2002).

Table 2.1: Physicochemical properties of the amino acid L-methionine (Barrett et al. 1985)

IUPAC IUB abbreviations Met; M
Molecular weight 149.22
Decomposition temperature 283
Water solubility (g/100g; 25C) 3.5
pK
COOH
2.28
pK
NH3
+
9.21
Isoelectric pH 5.74
Density (g/cm
3
) 1.34
Taste in aqueous solution tasty

O
O
S
C
5
H
11
NO
2
S
NH
2
H

Figure 2.2: Molecular structure and empirical formula of the amino acid L-methionine.

In metabolism L-methionine is a provider of methyl groups, e.g. for choline, adrenaline and
nucleic acid biosynthesis. It is an essentiell amino acid for humans and animals, so it is impor-
tant to get enough methionine through nutrition. The bioactive form is S-adenosyl-methionine
(SAM) (Kase et al. 1975; Schlenk et al. 1978), an important donor of methyl groups in organ-
isms.

In the biologic translation system the start of the protein biosynthesis always begins with the 3
nucleotides for L-methionine (AUG) as starting codon; but later a detachment of L-
methionine from the constructed poly peptide chain is often observed.

- 5 -
___________________________________________________________________________
2.2.1 DL-methionine production by chemical synthesis

The most important industrial method to produce DL-methionine is the chemical synthesis. It
is also the most economical until now. This is due to the fact that it is possible to use the ra-
cemic mixture for feeding additives because the human and animal organisms are able to con-
vert the D-form of methionine completely into the nutritive L-form by means of an oxidase
and transaminase (Hasegawa et al. 2005).

raw
material
crude oil
propene
acrolein
Evonik
process
methyl mercaptan hydrocyanic acid
natural gas
H
2
S CH
3
OH NH
3
air, N
2
H
2
O
DL-methionine
recirculation of:
CO
2
NH
3

Figure 2.3: Diagram of the chemical synthesis of DL-methionine by Evonik (former Degussa). Apart
from L-methionine Evonik produces also the substrates acrolein, methyl mercaptane and cyanic acid
(Pack 2004).

Figure 2.3 shows a description of the DL-methionine synthesis of the Evonik (former De-
gussa) company in Antwerpen. The most fundamental raw materials are crude oil and natural
gas, in addition basic chemicals like propene, methanol, ammonia and hydrogen sulphide are
used. The process has to pass several intermediate steps. At the end the product is obtained
through a reaction of acrolein, methyl mercaptan and hydrocyanic acid. The advantage of this
procedure is the relative high yield and the high rate of recirculation of carbon dioxide and
ammonia. The ressource consumption is comparatively low: per kilogram amino acid only
one kilogram crude oil has been consumed (Degussa 2004).

- 6 -
___________________________________________________________________________
2.2.2 L-methionine production by enzymatic synthesis

For pharmaceutical purposes an important industrial process to gain L-methionine is the acy-
lation of DL-methionine with the enzyme L-aminoacylase, which catalyses the hydrolysis
reaction from DL-acetylated amino acids to L-amino acids. The reaction is enantioselective;
so it is easy to separate the L-forms from the D-forms which are furthermore in the acetylated
condition. This process was first performed by the Japanese company Tanabe in 1969 to get
L-amino acids on an industrial scale (Leuchtenberger et al. 1988). Also, in recent research
there are studies about optimizing the conversion of the racemic mixture into pure L-amino
acids (May et al. 2000; Tokuyama et al. 1996).

Another enzymatic catalysis which was performed in the 1990s is the conversion of DL-
methyl-thio-ethyl-hydantoin with Arthrobacter aurescens DSM7330 to L-methionine
(Voelkel 1993; Stehr 1996), described in Figure 2.4.

HN NH
O
S
D - Methyl - thio - ethyl - hydantoin
H
3
C
HN NH
O
S
H
3
C
L - Methyl - thio - ethyl - hydantoin N - Carbamoyl - L - Methionine L- Methionine
H
2
O +
- H
2
O
COOH
HN NH
2
O
H
3
C
S
Racemase
+H
2
O
- CO
2
- NH
3
S
H
3
C
NH
2
COOH
N - Carbamoyl - L - Amino-
acid - Amidohydrolase
O
O
Hydantoinase

Figure 2.4: Description of the molecular mechanism for the conversion from DL-methyl-thio-ethyl-
hydantoin (MTEH) to L-methionine with Arthrobacter aurescens DSM7330 (modified according to
Stehr 1996).

The hydantoin hydrolysis is an economic synthesis for optical pure amino acids. Several hy-
dantoin derivatives as substrates can be received in cost-efficient ways by chemical synthesis.
A complete conversion by enzymatic catalysis is assured, there are no costs for derivatization
and racemization (Voelkel 1993).

Three enzymes are responsible for the formation of the L-amino acids, in Figure 2.4 shown
e.g. for the production of L-methionine (Stehr 1996):
- hydantoin-racemase: catalyzes the racemization from D- to L-hydantoins

- 7 -
___________________________________________________________________________
- hydantoinase: catalyzes the hydrolysis of the amide bond of the hydantoin circle
- N-carbamoyl-amino-acid-amidohydrolase: catalyzes the release of the amino acid by
separation of carbon dioxide and ammonia from carbamoyl intermediate product.

Important for this reaction is the activation of enzyme activity with the aid of an inducer, in
this case DL-N-3-methyl-5-indolyl-methyl-hydantoin (IMH). Without treatment of the cells
with this inducer the production rate of L-methionine would be ten times lower (Stehr 1996).

2.3 Corynebacterium glutamicum

2.3.1 General information

In 1957 Kinoshita et al. isolated a bacterial strain which was able to overproduce L-glutamic
acid in minimal media with glucose as carbon source and release the product in the extracellu-
lar environment. The isolated soil bacterium was named Corynebacterium glutamicum. In
taxonomic terms it belongs to the family of Corynebacteriaceae. Its cell wall formation is
very characteristic (gram positive), especially the existence of mycolic acids which surround
the entire cell as a structured layer (Eggeling et al. 2001). The wild type strains are mostly
able to grow aerobically on basic minimal media containing a carbon source like glucose,
phosphate, sulphate, ammonia and in addition biotin due to the fact that this bacterial species
is completely biotin deficient (Stansen 2005). Furthermore Corynebacterium glutamicum is
immobile and non-sporulating. Since the isolation in 1957 high amounts of L-glutamic acid
have been produced with new developed or advanced strains of this species.

The cell wall of Corynebacteriaceae has a special structure which is different from other
gram positive bacteria; this is shown in Figure 2.5. As you can see there, the peptidoglycan
layer is connected to the heteropolysaccharide arabinogalactan. The external mycolic acid
layer is linked again with the arabinogalactan (Eggeling et al. 2003). The influence of this cell
wall formation on the export of amino acids is not yet well defined, the research on this field
is described in chapter 2.3.5.

- 8 -
___________________________________________________________________________

Figure 2.5: Formation of the Corynebacterium glutamicum cell wall (right) in comparison to gram-
positive (left) and gram-negative bacteria (middle) (Eggeling et al. 2003).

Because of many experiences the scientists gained over the last decades about this organism
and its metabolic fluxes in context of amino acid production, Corynebacterium glutamicum
has become the most important bacterial strain for amino acid overproduction (Schmid 2002).
It has been observed that the regulatory system is much more simple than that of Escherichia
coli (Tosaka et al. 1986). There is information of the production of L-glutamic acid (e.g. Ki-
noshita et al. 1961), L-phenylalanine (e.g. Wartenberg 1989), L-lysine (e.g. Eggeling et al.
1999), L-valine (Blombach et al. 2007) and L-methionine (e.g. Kumar 2005) using strains of
Corynebacterium glutamicum.

Due to the historical importance of this particular field of research, a short description of
metabolic pathways for L-glutamic acid and L-lysine will be given in the following chapters
2.3.2 and 2.3.3.

- 9 -
___________________________________________________________________________
2.3.2 L-glutamic acid production with Corynebacterium glutamicum

The process to gain L-glutamic acid with Corynebacterium glutamicum by direct fermenta-
tion (Kinoshita et al. 1961) is very well investigated. Until today there is research about some
aspects of this topic (Stansen 2005). Key factors for the cultivation process in order to reach
high amounts of L-glutamic acid are the optimal concentration of biotin to influence and sup-
port cell growth and the secretion of the product in the extracellular environment (Clement et
al. 1986; Stansen 2005). Another important factor to prevent side reactions and by-products is
the oxygen supply. Under partially anaerobic conditions other / additional products like lactic
acid could be obtained (Kole et al. 1986).

Enzymes in Fig.2.6 (see numbers)

1: phosphoenolpyruvate carboxylase
2: pyruvate kinase
3: pyruvate carboxylase
4: pyruvate dehydrogenase
5: citrate synthase
6: aconitase
7: isocitrate dehydrogenase
8: L-glutamate dehydrogenase (GDH)
9: -ketoglutarate dehydrogenase (KDH)
10: isocitrate lyase
11: malate synthethase

Figure 2.6: Regulation of L-glutamic acid biosynthesis in Corynebacterium glutamicum (Leuchten-
berger 1996); straight lines represent feedback inhibition, dashed lines represent feedback repression.

The most important factor for L-glutamate overproduction is the activity of the enzymes GDH
and KDH (see Figure 2.6). In overproducers the conversion velocity of -ketoglutarate to L-
glutamic acid with GDH is 150 times higher than the side reaction of the substrate with KDH
which leads back to the citric acid cycle (Shiio et al. 1980). In Figure 2.6 the versatile regula-
tion mechanisms in biological pathways for L-glutamic acid (feedback inhibition and repres-
sion) are shown, the problems in modifying these metabolic fluxes in desired directions are
quite obvious due to the complexicity and various connections in these metabolic cycles.

- 10 -
___________________________________________________________________________
2.3.3 L-lysine production with Corynebacterium glutamicum

In contrast to L-glutamic acid production the wild types of Corynebacterium glutamicum are
not able to overproduce L-lysine, thus much research has been performed on modifying these
organisms with directed and random mutations to obtain auxotrophic and regulatory deficient
strains which are appropriate for L-lysine overproduction (Hilliger 1991).


1: phosphoenolpyruvate carboxylase
2: pyruvate kinase
3: pyruvate carboxylase
4: pyruvate dehydrogenase
5: citrate synthase
6: L-aspartate aminotransferase
7: aspartate kinase
8: aspartate-4-semialdehyde dehydrogen.
9: dihydropicolinate synthase
10: dihydropicolinate reductase
11: m-diaminopimelate dehydrogenase
12: m-diaminopimelate decarboxylase
13: homoserine dehydrogenase
14: homoserine kinase

Figure 2.7: Regulation of L-lysine biosynthesis in Corynebacterium glutamicum (Leuchtenberger
1996); straight lines represent feedback inhibition, dashed lines feedback repression, arrows with no
filling colour indicate activation.

Just like for L-methionine the production of L-lysine with Corynebacterium glutamicum is of
great importance as food additive for animal feeding. It is possible to accumulate L-lysine in
concentrations around 170 g/l in the fermentation process (Eggeling et al. 1999). In addition
L-lysine belongs to the same group of amino acids as L-methionine, the aspartate family; both
amino acids share intermediates for their own synthesis in metabolic pathways. The interme-
diate L-aspartate--semialdehyde (see Figure 2.7) is the branching point. L-Lysine overpro-
ducers possess strong activities of the enzyme dihydropicolinate synthase and reduced activi-
ties for homoserine dehydrogenase. If this ratio is contrary, the metabolic fluxes then lead to
the accumulation of the intermediate homoserine and subsequently to the formation of L-

- 11 -
___________________________________________________________________________
threonine and L-methionine. Also important is the inactivation of homoserine dehydrogenase
in order to direct the metabolic flux to the L-lysine biochemical pathway. Another key en-
zyme is the aspartate kinase which catalyzes the reaction from L-aspartate to L--aspartyl
phosphate. This enzyme is inhibited by both end products L-lysine and L-threonine, so it is
important to break this inhibition process (Wittmann et al. 2004; Eggeling et al. 1999).

Figure 2.8 presents a comparison between the industrial processes to gain high purificated L-
lysine hydrochloride on the one hand and the more economical process for the production of
Biolys (developed by the Evonik company). The Biolys process is much cheaper and gener-
ates less liquid and solid waste but there are no purification steps. It is allowed to use the
dried biomass of Corynebacterium glutamicum with high lysine concentrations as animal
food additive. If it would be possible to establish L-methionine overproduction (the second
important amino acid in animal food additives) with strains of C. glutamicum, in theory, the
same industrial process could be used.

NH
2
O
H
OH
H
2
N
Biolys
Corynebacterium glutamicum
C-source + N-source + salts + trace elements
fermentation fermentation
cell separation
ion exchange
evaporation
cristallisation
drying
L-Lysine-HCl
evaporation
spray drying
biomass
waste
ammonia
mother liquor

Figure 2.8: The Evonik (former Degussa) Biolys production scheme compared to the L-lysine-HCl
(lysine hydrochloride) scheme (Leuchtenberger et al. 2005).

- 12 -
___________________________________________________________________________
2.3.4 L-methionine production with Corynebacterium glutamicum

The first studies on microbial L-methionine production were performed in Japan in the 1970s
(Kase et al. 1974; 1975). They used strains of Corynebacterium glutamicum. Also, the Japa-
nese teams carried out the first research on the metabolic pathways of l-glutamic acid and l-
methionine as well as the screening methods to develop new strains of Corynebacterium glu-
tamicum (Shiio et al. 1980; Kase et al. 1975). The results on enzymes and precursors of L-
methionine biosynthesis showed differences between this pathway in Corynebacterium glu-
tamicum compared to other microorganisms. The greatest surprise was the development of the
intermediate product O-acetylhomoserine (Figure 2.9). In most organisms another form of
this product appears, O-succinylhomoserine. In Figure 2.9 it is shown that O-
acetylhomoserine is an intermediate product for Corynebacteria, whereas O-
succinylhomoserine is an intermediate product for Escherichia coli and Pseudomonas aerugi-
nosa. In contrast C. glutamicum and E.coli share the formation of cystathione in the next step
of biosynthesis, but in P. aeruginosa this precursor is omitted by direct catalysis to homocys-
teine, the last intermediate in L-methionine biosynthesis. Moreover, cystathione has the capa-
bility to use two alternative pathways leading to L-methionine synthesis: the trans-sulfuration
pathway and the direct sulfhydrylation (Hwang et al. 2002; Lee et al. 2003). As shown in Fig-
ure 2.9 and also in Figure 2.10 in the trans-sulfuration pathway cystathione is incorporated as
precursor and cysteine is metabolized. In the direct sulfhydrylation pathway cystathione is
bypassed and instead of cysteine inorganic sulfur is used for metabolization. Whereas most
organisms only have the ability to use one of these pathways, Corynebacterium glutamicum is
flexible enough to utilize both. Each pathway is regulated by independent enzymes and tightly
regulated by feedback repression.

- 13 -
___________________________________________________________________________

Figure 2.9: Regulation of L-methionine biosynthesis (Kovaleva 2007) in Corynebacterium glu-
tamicum (Cgl), Escherichia coli (Eco) and Pseudomonas aeruginosa (Pae); Cdp (Corynebacterium
diphteriae), Cef (Corynebacterium efficiens). The direct sulfhydrylation and trans-sulfuration path-
ways, which are alternatively used by different organisms, are indicated.

The regulation mechanisms and intermediates (feedback inhibition and repression) in Cory-
nebacterium are different compared to other organisms, e.g. Escherichia coli (Kovaleva et al.
2007). In spite of the various regulation systems shown in Figure 2.10, it is easier to perform
changed metabolic fluxes in this organisms than in others. This is due to the experiences re-
searchers obtained in working with Corynebacteria strains to produce other amino acids. Sev-
eral efforts have been made to characterize the L-methionine biosynthesis (Rckert et al.
2003; Kovaleva et al. 2007; Hwang et al. 2002).

In contrast to former investigations, where mostly the enzymes and activities were analyzed,
recent studies contain detailed descriptions of the genes, which are responsible for the enzyme
formation in L-methionine biosynthesis. Furthermore the possibilities to influence genes on
transcription level by silencing and enhancing signals and control the amount and formation
of key enzymes become increasingly dominant.

- 14 -
___________________________________________________________________________

1: aspartate kinase
2: aspartate semialdehyde
3: homoserine dehydrogenase
4: homoserine acetyltransferase
5: cystathione--synthase
6: cystathione--lyase
7: homocysteine methylase
8: methionine adenosyl transferase
9: O-acetylhomoserine sulfhydrylase
10: homoserine kinase

Figure 2.10: Description of the signaling, inhibition and repression phenomenons of L-methionine
biosynthesis in Corynebacterium glutamicum; dashed arrows indicate multiple steps. Straight and
dashed lines with circular dots at the ends indicate inhibition and repression, respectively. Pathways of
trans-sulfuration and direct sulfhydrylation are illustrated (Hwang et al. 2001).

The regulation mechanisms of L-methionine biosynthesis have been partially elucidated. The
separation from L-lysine biosynthesis takes place with the conversion from aspartate semial-
dehyde to homoserine which is catalyzed by homoserine dehydrogenase. In the next step of
synthesis the reaction to O-Acetylhomoserine means also the decoupling from the threonine
pathway. From this point there is no sharing of precursors for other amino acid synthesis
pathway any more. Figure 2.10 shows that the regulation of this catalysis by homoserine ace-
tyltransferase is the highest regulated step in this metabolic pathway. There are 5 feedback
inhibition and 2 feedback repression phenomenons influencing the enzyme activity or the
gene transcription. The end products L-methionine and SAM (S-adenosyl-methionine), the
bioactive form, have the capability to control the formation of several intermediates by inhibi-
tion or repression reactions.The chemical constitutional formulas of the biosynthesis of L-
methionine are shown in Figure 2.11 to give an overview of the structures of precursors and
reactions of chemical active groups important for the metabolic pathway of L-methionine.
Additionally, the responsible genes for enzyme formation are mentioned (Rckert et al.
2003).

- 15 -
___________________________________________________________________________
C
NH
3
+
H
COO
- CH
2
C
O
-
O
ATP
ADP
C
NH
3
+
H
COO
- CH
2
C
O
O
Aspartate kinase
(ask)
Aspartate
Aspartate phosphate
NADPH
NADP
+
+P
i
Aspartate semialdehyde
dehydrogenase (asd)
-2
O
3
P
C
NH
3
+
H
COO
-
CH
2
C
O
H
Aspartate semialdehyde
NADPH
NADP
Homoserine
dehydrogenase (hom)
C
NH
3
+
H
COO
-
CH
2
CH
2
OH
Homoserine
Acetyl - CoA
CoASH
C
NH
3
+
H
COO
-
CH
2
CH
2
O
Acetyl
Homoserine
acetyltransferase (metA)
O-Acetylhomoserine

- 16 -
___________________________________________________________________________
O-Acetylhomoserine (see last page)
Cystein
Acetat
Cystathione
y-synthase (metB)
C
NH
3
+
H
COO
-
CH
2
CH
2
C
NH
3
+
COO
-
H
S
Cystathione
- Pyruvat
- NH
4
+
C
NH
3
+
H
COO
-
CH
2 HS
CH
2
CH
2
Cystathione
-lyase (aecD)
Homocysteine
N
5
- Methyl - THF
THF
C
NH
3
+
H
COO
-
CH
2
S CH
2
Methionine
Homocysteine
methylase (metE/metH)
glyA
(metF)
direct
sulfhydrylation
pathway
O-acetylhomoserine
sulfhydrylase (metY)

Figure 2.11: Pathway for the intermediates of L-methionine biosynthesis in Corynebacterium glu-
tamicum in structural formulas; product names are written on the right, enzyme names on the left side,
genes in brackets (modified after Voet et al. 2002 and Rckert et al. 2003).

- 17 -
___________________________________________________________________________
Obvious in this context in Figure 2.11 is the existence of 2 genes (metE/metH) for the enzyme
homocysteine methylase. MetE is a vitamin B
12
independent, metH a vitamin B
12
dependent
methylase (Rckert et al. 2003).

In first studies on L-methionine production (Kase et al. 1974; 1975; Banik et al. 1974) the
Corynebacteria strains were improved by random mutation and selection with methionine
analogues / antimetabolites like ethionine. The strains were capable to produce approximately
2 g/l after 3 days.

In the 1990s research groups in India worked with auxotrophic mutants of Corynebacteria
and other microbial strains on L-methionine overproduction (Mondal et al.1990; 1994; 1996).
The strain improvement was also achieved by random mutations and L-methionine antime-
tabolites. L-methionine concentrations over 10 g/l in shake flask cultures could be reported.

In the last ten years more literature and information became available on L-methionine pro-
duction with improved strains achieved by random mutation of Corynebacterium glutamicum
(Sharma et al. 2001; Kumar et al. 2003). Concentrations of around 2.3 g/l of L-methionine
were obtained according to these sources. In 2005 a comprehensive literature review of L-
methionine producing strains was published (Gomes et al. 2005).

These and other studies also provide much information about cultivation conditions. The me-
dia composition for Corynebacterium glutamicum strains to gain optimal growth (Liebl et al.
1989) and effects of parameters like oxygen supply and pH to achieve higher production rates
were described widely in the literature (e.g. Kumar et al. 2005).

The first study about attempts in genetic engineering of Corynebacterium glutamicum me-
tabolism to overproduce L-methionine did not show high overproduction rates of L-
methionine (Mampel et al. 2005), which means that a real break-through has not been
achieved so far.

- 18 -
___________________________________________________________________________
2.3.5 Amino acid excretion mechanisms in Corynebacterium glu-
tamicum

The excretion of amino acids is an important factor for overproducing microorganisms. If the
product is accumulated intracellularly, an additional step in down-stream processing is neces-
sary, the cell disruption. For this reason the process would be much more expensive in indus-
trial categories compared to product secretion in the extracellular environment. A high
amount of accumulated substance in the cytosol has also other negative side effects. The pro-
duction rate could be slowing down by cell signaling in order to avoid intracellular destruc-
tion.

For Corynebacterium glutamicum, which is a biotin deficient strain, the amount of this vita-
min in minimal media is an important factor to get optimal excretion (Clement et al. 1986).
According to this data it is easier for the organism to secrete L-glutamic acid while the biotin
concentrations are in starvation conditions due to the incomplete formation of cell wall mate-
rials. In addition a temperature induced release of L-glutamic acid and its precise characteri-
zation has been reported recently (Stansen 2005). For L-glutamic acid there are reports about
the active transport through the cell wall with a specific carrier system which has not been
fully characterized and understood until now (Nakamura et al. 2007). Also there are charac-
terizations of the lysine exporter LysE and the threonine exporter ThrE (Eggeling et al.
2003). For L-methionine connections between export and biotin concentration were observed
(Mondal et al 1994).

Mechanisms for the general possibilities of the L-methionine export were described since the
90s (Krmer 1993). The export of amino acids is based in most cases on a secondary transport
system. New results about the export system for L-methionine were published in 2001 and
2005 (Eggeling et al. 2001; Trtschel et al. 2005). According to this data the L-methionine
export systems are tightly regulated. The overexpression of distinct membrane proteins is
linked with increased intracellular concentrations of methionine. One responsible gene, BrnF,
coding for the larger subunit of BrnFE, has been identified as being responsible for one
methionine export system (and furthermore also for isoleucine export). Additionally, strong
indications for the presence of at least one other export system, presumably with low affinity
but high capacity, exist due to the investigations in the works of Eggeling (2001) and
Trtschel (2005).

- 19 -
___________________________________________________________________________
2.4 Biochemical amino acid degradation

Apart from sugars, amino acids are common substrates for microbial cell growth. Sources of
amino acids, like peptones, tryptones and casamino acids are used in laboratory media. The
degradation of individual amino acids depends on the availability of other energy sources and
on the C/N ratio.

The first step in biochemical amino acid degradation is, in general, the removal of the -
amino acid group. The reaction products are 2-oxo acids. There are different biochemical
ways to perform this reaction (Lengeler et al. 1999). They are, as follows:

- the oxidative deamination
- the transamination
- the -elimination

After the -amino group removal further degradation steps take place according to the back-
bones of the 2-oxo acids (aliphatic, heterocyclic, aromatic compounds).

The biochemical degradation products of amino acids are normally intermediates of the citric
acid cycle or precursors of these intermediates. They are converted to carbon dioxide and wa-
ter or will be applied in the gluconeogenesis. The degradation of methionine and other amino
acids which are important in this thesis is shown in Figure 2.12. The amino acids can be di-
vided in two groups of catabolic degradation, the glucogenic and the ketogenic amino acids.
The ketogenic amino acids are converted to acetyl-CoA or acetoacetate. Afterwards they are
transformed to fatty acids or ketone bodies. Lysine, leucine and threonine belong to this group
of amino acids which can be degraded only this way. The degradation products of these three
amino acids cannot be used for gluconeogenesis. Tryptophane, tyrosine, iso-leucine and
phenylalanine are amino acids which can be metabolized on both biochemical ways for deg-
radation, the ketogenic and the glucogenic one. The other amino acid groups, shown in Figure
2.12, are degraded through the glucogenic ways. This means that their degradation products
are available for gluconeogenesis. Glucose precursors like pyruvate, -ketoglutarate and suc-
cinyl-CoA are synthesized on these biochemical pathways. Methionine is converted to suc-
cinyl-CoA, glutamic acid to -ketoglutarate and alanine as well as glycine are degraded
through transamination to pyruvate and following acetyl-CoA. The biochemical degradation

- 20 -
___________________________________________________________________________
is an important explanation approach to discuss increased and decreased amino acid concen-
trations under nutrient deficient condition phases in shake flask and bioreactor cultivations.
Under those conditions the microorganisms try to find possibilities to maintain the metabo-
lism by degradation or conversion of available substances, like e.g. the conversion of amino
acids (Voet et al. 2002).

Figure 2.12: Biochemical amino acid degradation to intermediates of the citric acid cycle. The gluco-
genic degradation is coloured in green, the ketogenic one in red / violet (Voet et al. 2002).

- 21 -
___________________________________________________________________________
2.5 Random mutagenesis and following selection

Random mutation is a basis method in biochemistry to improve the abilities of microorgan-
isms. Mutation is performed by the use of highly potent chemical mutagens for bacteria like
NTG (N-methyl-N-nitro-nitrosoguanidine) and EMS (ethyl-methanesulfonate) or ultraviolet
(UV) radiation. NTG induces primarily base transition mutations of the GC AT type, al-
though ATGC transitions, transversions, and even frameshifts appeared at low frequencies
(Gerhardt et al. 1981). EMS works in similar ways, both NTG and EMS are also dangerous
carcinogens! The ultraviolet radiation of microorganisms does not have the same potential in
mutation compared to EMS and NTG; working with this method is, however, much less dan-
gerous. Therefore, the handling of UV radiation in screening systems is far easier and the
process is much more ecological because due to the natural occurrence of UV radiation an
environment which influences all living organisms on earth. Usually UV rays around 250 nm
are deployed in experiments.

Shake flask culture
1.Mutation

UV
Minimal media
Agar plate
Agar plate
Minimal media with
anti-metabolites
(L-Methionine analogues)
2. Stamp techniques
(Replica plating)
Colonies resistant
to anti-metabolites
R Re ea as so on n: : Overproduction of the real metabolite
(in this case L Methionine) due to a regulation defect
3. Mutant isolation
4. Cultivation
5. Production of the real
metabolite

Figure 2.13: Concept for random mutation and the succeeding selection method to get overproducers
for the amino acid L-methionine.

The selection method is crucial because the multitude of obtained microorganisms has to be
reduced for analysis. In Figure 2.13 the principle method for the random mutation and the
following selection step is documented for the case of L-methionine. In this example the se-
lection is performed by replica plating (a wide-spread microbiological method) on agar plates

- 22 -
___________________________________________________________________________
with minimal media and, in addition, L-methionine analogues / anti-metabolites. These anti-
metabolites are almost molecular structural analogues to L-methionine. In theory all microor-
ganisms with functional feedback and inhibition signaling systems do not produce L-
methionine in the presence of high L-methionine concentrations in media but metabolize the
available supply. If the microorganisms mix up methionine with the L-methionine analogues
and deactivate their production pathways, then microbial death is unevitable. It is not possible
to utilize analogues of amino acids for metabolism. On the other hand mutant strains derived
from UV radiation with deactivated signaling sytems for distinct feedback inhibiton and re-
pression systems are able to produce L-methionine in high amounts. These mutants which are
resistant to methionine analogues have altered and deregulated enzymes that are not sensitive
to feedback inhibition or repression due to mutations in structural important genes for this
signalling system (Kumar et al. 2005).

There are various structural analogues of L-methionine described in the literature like ethion-
ine, norleucine, methyl methionine, methionine sulfoxide, methionine methylsulfonium chlo-
ride and others. The difference of the constitutional formula of L-methionine compared to the
structural analogues L-ethionine and L-norleucine is shown in Figure 2.14. L-Ethionine has a
CH
2
-group (methylene) more as L-methionine, L-norleucine does not possess a sulphur com-
ponent. The usage of ethionine is mentioned as the most effective substance (Mondal et al.
1994. Kumar et al. 2003). Nevertheless the development of fast and effective analysis systems
is absolutely required to enhance the chances to obtain mutant strains with improved proper-
ties.
O
OH
NH
2
H
C
6
H
13
NO
2
S
O
OH
NH
2
H
C
6
H
13
NO
2
S
S
O
OH
NH
2
H
C
5
H
11
NO
2
S
L-Norleucine
L-Ethionine
L-Methionine

Figure 2.14: Comparison of the chemical structure of the amino acid L-methionine and the structural
analogues / anti-metabolites L-ethionine and L-norleucine.

- 23 -
___________________________________________________________________________
2.6 Amino Acid Analysis

The introduction and refinement in analysis of amino acids, especially L-methionine, is an-
other important point to get reproducible results, particularly with regard to the huge amounts
of samples achieved by the screening.

Various methods for the analysis of amino acids are mentioned in the literature. In former
investigations it was common to separate amino acid mixtures with paper chromatography,
thin layer chromatography (TLC) or ion exchange chromatography (IEC). Afterwards, the
analysis was performed by ninhydrin reaction (Kase et al. 1974). In presence of ninhydrine
the heating of amino acids leads to an oxidative desaminination und at the same time a decar-
boxylation of the amino acids. This results in formation of carbon dioxide and an aldehyde
which is one C-atom shortened. For quantitative acquisition of data blue violet dye (Ruhe-
manns violet), formed in a following reaction, has been used (Stahl 1967). Today the detec-
tion systems for amino acids are high pressure liquid / performance chromatography (HPLC)
and gas chromatography (GC) (Husek et al. 2001).

Many studies are available about a derivatization of amino acids before separation in order to
make the following detection easier. On the basis of Edmans degradation of proteins (Voet et
al. 2002) with phenylisothiocyanate (PITC) to obtain the amino acid sequence, the analysis of
amino acid mixtures by derivatization with PITC and detection with HPLC is widely de-
scribed (e.g. Heinrikson et al. 1984). One of the state-of-the-art techniques is the derivatiza-
tion with ortho-phtaldialdehyde (OPA). This reaction is distinguished by high sensitivity but a
problem is the missing reaction with secondary amines. An alternative is e.g. the reaction with
6-amino-quinolyl-N-hydroxy-succinimidyl-carbamate (ACQ). (Naidanow et al. 2005). Today
a broad range of derivatization agents for amino acid detection are available. It is common to
separate the amino acids with reversed phase chromatography, which means that the station-
ary phase (column solid material) is unpolar (e.g. by reaction of silica gel with silanes) and
the mobile phase (the eluent) is polar (Heinrikson et al. 1984 + Husek et al. 2001). Special
chiral colums for separation and detection of enantiomeric forms of amino acids in HPLC and
GC are developed. The combination sytem of GC / HPLC with mass spectrometry (MS) al-
lows to get a molecular fingerprint of each chemical compound which is an important tool for
reliable identification of substances.

- 24 -
___________________________________________________________________________
Due to problems with derivatization and in order to accelerate the analysis, attempts to detect
amino acids in unmodified conditions with commercial liquid detectors have been made
(Petritis et al. 2002). In comparison with several detector types, the chemiluminescent nitro-
gen detection (CLND) and the tandem mass spectrometry (MS) show high sensitivity and
specificity.

- 25 -
Materials and methods
_________________________________________________________________________
3 Materials and methods

3.1 Microbial strains

Arthrobacter aurescens DSM7330
This bacteria also belongs to the group of coryne-form bacteria. It was isolated from soil
samples in screening experiments to convert aryl hydantoins (Voelkel 1993). The physio-
logical and morphological properties are the positive gram staining, strictly aerobic, pleo-
morphic growth and the amotility. The organism is nonsporulating. In this research the
strain was used to demonstrate the efficiency of obtaining L-methionine in a biocatalysis
experiment. It is the target to show the broad spectrum of possibilities to gain amino acids
like L-methionine.

Corynebacterium glutamicum DSM20300 / ATCC13032
This is the most common C. glutamicum strain for investigations. In 1957 the soil bacte-
rium was isolated under the name Micrococcus glutamicus (e.g. Kinoshita et al. 1957). The
ability to produce L-glutamic acid was described widely in the literature (Kinoshita et al.
1961). It is the base material for all research on amino acid overproduction and is available
at all famous cell culture collections like DSM (Deutsche Sammlung fr Mikroorganismen
und Zellkulturen, Germany) and ATCC (American type culture collection). In this study
the strain has been used to reproduce the L-glutamic acid production and in random
mutagenesis in order to get new overproducers of L-methionine.

Corynebacterium glutamicum ATCC21608
This strain is commercially available at the ATCC. According to information (Nakayama
et al. 1973) the microorganism might be able to produce L-methionine. It was obtained by
random mutation with methyl methionine as structural analogue/anti-metabolite. The base
material for the experiments was the strain C. glutamicum DSM20300 / ATCC13032.

- 26 -
_________________________________________________________________________
Corynebacterium glutamicum KY10574
This strain was also received by random mutation of C. glutamicum DSM20300 /
ATCC13032 to improve the L-methionine production abilities. It was directly delivered by
the Japanese company Kyowa Hakko Kogyo Ltd, Tokyo. It is the most important strain for
this study, several experiments described in this work were carried out with these bacteria
in order to optimize the production rates and to get information about adequate physico-
chemical and biochemical parameters for cultivation. In the picture in Figure 3.1 (recorded
with an electron microscop) the rod-shaped form of the bacterial species Corynebacterium
glutamicum is shown in many examples. The coryne-form of the cells is not that clear to
see due to the fact, that Corynebacterium is a pleiomorphic (form variability) species
(Lengeler et al. 1999). Packages of several cells in parallel arrangement (palisades) can
also be observed.

5 m

Figure 3.1: Electron microscopic picture of Corynebacterium glutamicum cells, recorded at the
Forschungszentrum Jlich, Institute of Biotechnology (IBT).

All bacteria of the species Corynebacterium glutamicum possess the GRAS status (gener-
ally regarded as safe). For general information about universal Corynebacterium glu-
tamicum characteristics, see chapter 2.3.1.

- 27 -
_________________________________________________________________________
3.2 Media and biotransformation experiments with Ar-
throbacter aurescens DSM7330

The experiments with this strain were specifically described in several PhD - studies from
the 1990s which had been undertaken at the Institute for Biochemistry and Biotechnology,
TU Braunschweig (Voelkel 1993; Stehr 1996). Therefore, access to these microbial mate-
rial was easy to get. Thus, there will be only a short summary of media composition and
the way to perform the biocatalysis experiments in this research (for further information
please have a look in the PhD theses from Voelkel and Stehr).

The following information about media compositions and biocatalysis are based on the
studies from Stehr (1996).

The first step of the experiments was the microbial growth of Arthrobacter aurescens to
gain biomass and to activate the enzyme acitivities using the inducer DL-N-3-methyl-5-
indolyl-methyl-hydantoin (IMH). The biomass was centrifuged and afterwards employed
in the biotransformation investigations.

The biotransformation was performed in 100 ml shake flasks (each with 2 baffles). The
vessels were pre-filled with 9 ml substrate solution (0.1 molar glycine / NaOH buffer, 5.7
mM DL-MTEH (methyl-thio-ethyl-hydantoin)). After that 1 ml of wet biomass in a con-
centration of 50 g/l was added. The biotransformation reaction was carried out in a water
bath vessel for 24 h at 37C. The initial pH of 8.5 was controlled and adjusted correctly in
the first 5 hours. The samples were centrifuged (Heraeus Christ Biofuge A) for 15 min at
13,000 rpm and stored at 4C in the refrigerator. The supernatant was analysed with HPLC
(high pressure liquid chromatography).

- 28 -
_________________________________________________________________________
3.3 Corynebacterium glutamicum media

3.3.1 Conservation media

The conservation of the Corynebacteria strain material was carried out using a complex
BY (bouillon yeast extract) medium (see table 3.1):
After inoculation 50 ml BY medium in a 250 ml shake flask (2 baffles) was cultivated for
one day (parameters: 30C, 120 rpm). Afterwards, for short time conservation BY agar
media plates were inoculated with 50 l and also cultivated for one day. These agar plates
were stored at 4C in the refrigerator. They form the basis material for all experiments. The
long time conservation was carried out with glycerine stocks. 1 ml of a bacterial sample
was mixed with 1 ml glycerine in plastic tubes and stored at -20C.

Table 3.1: Composition of complex BY (agar) medium for strain conservation.

20 g standard-I-nutrition bouillon
5 g yeast extract
22 g agar-agar (only for solid plate media)
pH 7.2 adjusted with 6N NaOH
fill up to 1,000 ml with distilled H
2
O; sterilization

- 29 -
_________________________________________________________________________
3.3.2 Media for Corynebacterium glutamicum DSM20300

There were two minimal media in use for the strain Corynebacterium glutamicum
DSM20300, one for the screening after L-methionine overproducers, one for the reproduc-
tion of L-glutamic acid production. The minimal medium for the screening system was
approximated to L-lysine production media CGXII with some modifications (Keilhauer et
al. 1993; Deutenberg 2003).

Table 3.2: Composition of screening minimal medium SMM

Solution 1: basic components
20 g (NH
4
)
2
SO
4
10 g (NH
4
)
2
S
2
O
3
5 g urea

1 g KH
2
PO
4
2 g K
2
HPO
4
0.25 g MgSO
4
7 H
2
O
0.3 mg 3,4 - dihydroxy benzoic acid
fill up to 894 ml with distilled H
2
O
sterilisation

Solution 2: glucose
50 g glucose
2
O
sterilisation

Solution 3: mineral elements
200 mg CaCl
2
200 mg FeSO
4
7 H
2
O
200 mg MnSO
4
H
2
O
20 mg ZnSO
4
7 H
2
O
6 mg 3,4 - dihydroxy benzoic acid
1.2 mg CuSO
4
200 g NiCl
2
6 H
2
O
2
O
sterile filtration

Solution 4: vitamin stock solution
200 mg biotin
200 mg thiamin hydrochloride
2
O
sterile filtration

- 30 -
_________________________________________________________________________
Composition of SMM medium:

894 ml solution 1, 100 ml solution 2, 5 ml solution 3 and 1 ml solution 4 were pooled. The
medium SMM was used in several experiments in different scales. For the screening ex-
periments in 96 well plates amounts of 250 l medium per well were used. The medium
volume in the shake flask and bioreactor scales contained up to 3.5 litres. The temperature
in all experiments was 30C, in the shake flask experiments shaking rates of 120 rpm were
used.

For the production of L-glutamic acid the media were made according to instructions from
the original patent sources (Kinoshita et al. 1961).

Table 3.3: Composition of media for L-glutamic acid production

Complex preculture for L-glutamic acid production CP
10 g peptone, tryptic
10 g yeast extract
5 g glucose
5 g NaCl
2
O
pH 7.2 adjusted with 2N NaOH and 2 molar HCl
sterilisation

Minimal production medium for L-glutamic acid production MPM
5 g urea
0.1 g MgSO
4
7 h
2
O
0.5 g K
2
HPO
4
0.005 g FeCl
3
6 H
2
O
2
O
pH 7.5 adjusted with 2N NaOH and 2 molar HCl
sterilisation
addition of separately sterilised glucose solution (50g / 100 ml)

The precultivation was performed in a 500 ml shake flask with 2 baffles with 100 ml of CP
medium at a temperature of 30C and 100 rpm for 17 h. The inoculation was done directly
with a loop of cell material from agar plates (BY medium, see chapter 3.3.1). The main
culture was inoculated with 2 ml of the incubated preculture on 100 ml of minimal produc-
tion medium MPM (see Table 3.3). Afterwards, the cultivation was performed under the
following conditions: 144 h, 30C, 100 rpm, 102 ml working volume in a 500 ml shake
flask (2 baffles).
- 31 -
_________________________________________________________________________
3.3.3 Media for Corynebacterium glutamicum ATCC21608

The strain Corynebacterium glutamicum ATCC21608 was investigated using the same
media as for the strain Corynebacterium glutamicum KY10574. For further information
about these media (KH2 / KH1 / F1) see chapter 3.3.4.

3.3.4 Media for Corynebacterium glutamicum KY10574

The first cultivation instructions (media KH1 und KH2, see Tables 3.4 and 3.5) for the
strain C. glutamicum KY10574 were obtained from Kyowa Hakko Kogyo Ltd. The com-
plex media KH1 was directly inoculated from an agar plate, after 24 h cultivation time (pa-
rameters: 30C, 120 rpm, 25 ml medium in a 250 ml shake flask (2 baffles) the production
media KH2 was inoculated in a 1:11 ratio.

Table 3.4: Composition of complex seed medium KH1(Kyowa Hakko)

10 g peptone, tryptic

10 g yeast extract
2.5 g NaCl
2
O; sterilization
addition of separately sterilised glucose solution (20 g/200 ml)

The production medium KH2 (see Table 3.5) is cloudy due to the very unsoluble CaCO
3

component. The cultivations were performed in 11 ml scale in a 250 ml shake flask with 2
baffles (1 ml KH1 + 10 ml KH2) and 110 ml scale in a 1 L shake flask (10 ml KH1 + 100
ml KH2) for 4 days. The experiments were carried out at 30C and 120 rpm.

- 32 -
_________________________________________________________________________
Table 3.5: Composition of production minimal medium KH2 (Kyowa Hakko)

20 g (NH
4
)
2
SO
4
0.5 g K
2
HPO
4
0.5 g KH
2
PO
4
1 g MgSO
4
7 H
2
O
100 mg FeSO
4
7 H
2
O
100 mg MnSO
4
4 H
2
O
20 g CaCO
3
2
O; sterilization
addition of separately sterilised glucose solution (100 g/300 ml)
sterile filtration: 1 ml biotin (from 100 mg/l stock solution)

In cooperation with the vTi Braunschweig (Johann Heinrich von Thnen Institute, Institute
of Agricultural Technology) and after an extensive literature search concerning Coryne-
bacterium glutamicum a minimal medium (F1 medium, see Table 3.6) has been developed
in order to get bacterial growth without the influence of complex media components. The
medium was produced through mixing of several solutions (see Table 3.6).

The mixing ratios of the liquids were as follows: 200 ml solution A, 100 ml solution B, 5
ml stock solution C, 300 ml solution D, 5 ml stock solution E, 2 ml stock solution F, 350
ml solution G, 5 ml solution H, 10 ml stock solution I, 10 ml stock solution J. Afterwards,
the liquid amounted to 987 ml.

5 ml F1 medium were taken out as aliquote to adjust pH to 7.3 with NaOH, the appropriate
amount of NaOH was added to the medium, last of all the medium was filled up with H
2
O
to 1000 ml.

In order to get information about influence of carbon source on amino acid production (es-
pecially L-methionine), the glucose concentration was altered in the following concentra-
tion ranges between 20 and 100 g/l.

As well as for the KH-media described before the cultivations were performed in compa-
rable scales. In a 250 ml shake flask with 3 baffles a working volume of 25 ml was used, in
a 1 L shake flask 100 ml working volume was utilized. The general parameters were also
30C for the temperature; the shaking speed varied between 120 and 210 rpm.

- 33 -
_________________________________________________________________________
Table 3.6: Composition of F1 minimal medium

Solution A 20 g glucose

2
O; sterilization

Solution B 30 g KH
2
PO
4
60 g K
2
HPO
4
5 g NaCl
2
O; sterilization

Solution C 6 g MgCl
2
2.5 g CaCl
2
2 H
2
O
2
O; sterile filtration

Solution D 10 g (NH
4
)
2
SO
4
2
O; sterilization

Solution E 4 g FeSO
4
7 H
2
O

50 g MgSO
4
7 H
2
O

1 g MnSO
4
H
2
O
1 g NaCl
fill up to 1,000 ml with 0.1 molar HCl; sterile filtration

Solution F 100 mg Na
2
BB
4
O
7
10 H
2
O
50 mg (NH
4
)
6
MoO
24
4 H
2
O

800 mg ZnSO
4
7 H
2
O
270 mg CuSO
4
5 H
2
O
7.5 mg MnCl
2
4 H
2
O
870 mg FeCl
3
6 H
2
O
10 mg NiCl
2
6 H
2
O
fill up to 1,000 ml with 0.05 molar HCl; sterile filtration

Solution G 5 g urea

2
O; sterilization

Solution H 3 g 3,4 - dihydroxy benzoic acid

2

Solution I 100 mg biotin

2

Solution J 20 mg thiamin hydrochloride

20 mg cyanocobalamin

2

- 34 -
_________________________________________________________________________
3.4 Analysis techniques

3.4.1 Bio dry mass and optical density

The bio dry mass was measured using metal tubes with 15 ml liquid samples and small
plastic reaction vials (Eppendorff caps) with 1 ml liquid samples. The small vials were
used due to the limitation of the amount of liquid in small scale cultivations in shake
flasks. The measurement of a single sample was always performed twice to sure the accu-
racy. The samples were centrifuged for 15 minutes at 12,000 rpm (Heraeus Christ Biofuge
A for 1.5 ml vials; Heraeus Sepatech Biofuge 17RS for metal tubes), the supernatants were
discarded. The vessels were stored in a heater (100C) for 24 h, afterwards the bio dry
mass in the vessels was measured against the empty weight. The samples with the nearly
unsoluble CaCO
3
medium component were measured by giving 200 l of 4 molar hydro-
chloric acid to a volume of 800 l sample liquid in order to dissolve the CaCO
3
compo-
nent.

The optical density was measured photometrically at a wavelength of 546 nm (photometer
type: Ultrospec 3000; Pharmacia Biotech). The samples were diluted in a ratio of 1:100 in
order to keep the absorption level in the linear area. The blank sample was the sterilised
medium which was also diluted in a ratio of 1:100.

3.4.1.1 Connection between the BDM and OD
546
values

The connection between the measurement of bio dry mass (BDM) and the optical density
at 546 nm (OD
546
) in F1 minimal medium is presented in Figure 3.2. Both measurements
are indicators for cell growth of the C. glutamicum microorganisms. The values which
were obtained from the measurement of both parameters in the shake flask cultivations as
described in chapter 4.4.3 were plotted in this graph. The development of the curves for
both parameters was nearly parallel in almost all cases shown in this chapter (see Figures
4.19 4.24).

- 35 -
_________________________________________________________________________
OD
546

0 10 20 30 40 50 60 70 80 90 100
B
D
M

(
g
/
l
)
0
2
4
6
8
10
12
14
16
18
20

Figure 3.2: Connection between the parameters for the cell growth of C.glutamicum KY10574 in
F1 minimal medium, the plotted values were obtained from the shake flask cultivations described
in chapter 4.4.3.

3.4.2 pH-measurement

The pH measurement was done potentiometrically with a pH-electrode type Schott CG
802. For calibration 2 control solutions (pH 4.01 and 7.00) were used.

3.4.3 Glucose measurement

The glucose measurement was performed with an enzymatic photometric test system (pho-
tometer type: Ultrospec 3000; Pharmacia Biotech) for D-glucose, by the companies Boe-
hringer Mannheim / R-Biopharm. The detection based on UV-absorption at 365 nm.

- 36 -
_________________________________________________________________________
The test system contains three liquid chemicals in bottles:

- The first solution (now named solution 1) consists of 7.2 g powder mixture, which
has to be dissolved in 45 ml H
2
O. According to the producer this solution 1 con-
tains triethanolamine buffer (pH approx. 7.6), NADP (approx. 110 mg), ATP
(approx. 260 mg) and magnesium sulfate.
- The second liquid chemical is a suspension (now named suspension 2) consisting
of the enzymes hexokinase (approx. 320 Units) and glucose-6-phosphate dehydro-
genase (approx. 160 Units).
- Solution 3 contains a standard solution of D-glucose for assay control purposes.

The reagents were directly given and mixed in plastic cuvettes suitable for photometric
purposes, the final volume was 3.020 ml. The reaction has to be carried out at a tempera-
ture of 20 25C. A schematic plan of the procedure is presented in Table 3.7.

Table 3.7: Schematic plan of the substance mixing for D-glucose measurement (Enzymatic test
system Boehringer Mannheim / R-Biopharm)

pipette into cuvette blank sample
solution 1 1,000 ml 1,000 ml
sample - 0.1 ml
H
2
O, dest. 2,000 ml 1,900 ml
mix and read absorbances of the solutions (A
1
) after approx. 3 min.
the enzymatic reaction is started by addition of:
suspension 2 0.02 ml 0.02 ml
mix and read absorbances (A
2
) after the reaction has stopped (approx. 10 15 min)

The reaction mechanism of this D-glucose measurement system is based on two enzymatic
reactions:
D-glucose +ATP
hexokinase
glucose-6-phosphate +ADP
glucose-6-phosphate +NADP
+

D-gluconate-6-phosphate +NADPH
+
+H
+
G6P-DH

- 37 -
_________________________________________________________________________
The amount of NADPH
+
formed in this reaction is stoichiometric to the amount of D-
glucose. The increase in NADPH is measured by means of its light absorbance at 365 nm.

The general equation to calculate the concentration is:

e glu D
A
v d
MW V
c
cos
1000

=

c = concentration of D-glucose (g/l), V = final volume (ml), v = sample volume (ml), MW
= molecular weight of the substance to be assayed (g/mol, 180.16 g/mol for D-glucose); d
= light path (cm), = extinction coefficient of NADPH at 365 nm =3.5 (L mmol cm
-1
),
A = Absorbance differences (A
2
-A
1.
see Table 3.7)

For D-glucose:

e glu D e glu D
A A
v d
c
cos cos
441 . 5
1000
16 . 180 020 . 3

=

=

If samples were diluted, the result had to be multiplied by the dilution factor (for further
information about the details see reference material of Boehringer Mannheim / R-
Biopharm).

- 38 -
_________________________________________________________________________
3.4.4 Thin layer chromatography (TLC)

The thin layer chromatography (TLC) is a quick and cheap method to detect amino acids in
a qualitative way. The samples dissolve in an appropriate solvent (mobile phase) and are
drawn up on the solid stationary phase via capillary action. The amino acids are separated
according to their polarities and the strength of their interactions with the stationary phase.
The detection is based on a dye producing reaction of the amino acids with ninhydrin.

For the experiments in this work a thin layer of silica gel (Silica gel 60 F
254.
Merck KGaA)
immobilised on aluminium sheets was used as stationary phase. The samples were placed
on the aluminium sheets in an amount of 5 l with a micro pipette. After drying, the alu-
minium sheets were placed in a closed glass chamber with the mobile phase which covered
the chamber ground (fill level: 1 cm). The mobile phase was a solution mixed of n-butanol
/ glacial acetic acid / H
2
O in a ratio of 12:3:5 (v/v/v). It was important to prevent immedi-
ate contact between the solvents (mobile phase) and the samples because the amino acid
spots then were very clear defined. After chromatography the aluminium sheet was dried
with a heater, then the surface of the sheet was sprayed with ninhydrin reagent. Finally the
sheet was dried once again on the heater.

The thin layer chromatography of the biotransformation samples was performed using a
chiral TLC plate (Macherey-Nagel company, chiral plate coated with Cu
2+
and chiral re-
agent), which was capable of separating DL enantiomeric product forms. The mobile phase
in this case consisted of a mixture of methanol / H
2
O / acetonitrile in a ratio of 50:50:200
(v/v/v). The other working stages and the ninhydrin detection were carried out in the same
manner as for the other TLC experiments.

- 39 -
_________________________________________________________________________

dye (Ruhemanns purple)

Figure 3.3: Reaction of ninhydrin with amino acids (Stahl 1967)

As shown in Figure 3.3 the reaction of ninhydrin with amino acids results in the separation
of water and the formation of a Schiff base. The following reactions are the decarboxyla-
tion of the amino acid and in the next step an oxidative desamination. Carbon dioxide and
a one C-atom shortened aldehyde are formed. This aldehyde reacts again with ninhydrin to
the dye (Ruhemanns purple). The formation of the dye is proportional to the concentration
of the amino acids. The method is very sensitive for amino acids, as shown in Figure 3.4.

- 40 -
_________________________________________________________________________

1 g/l 0.5 g/l 0.2 g/l 0.1 g/l

Figure 3.4: Qualitative thin layer chromatography of L-methionine samples in concentration
ranges from 0.1 g/l to 1 g/l; mobile phase is n-butanol / glacial acetic acid /H
2
O (12:3:5); stationary
phase is immobilised on aluminium sheet: Silica gel 60 F
254
; ninhydrin detection reaction.

3.4.5 High pressure / performance liquid chromatography (HPLC)

High pressure / performance liquid chromatography (HPLC) was used as a method to de-
termine the concentrations of the biotransformation experiments with Arthrobacter aures-
cens DSM7330. The method was developed in the 1980s (Battilotti et al. 1988) and opti-
mised in the 1990s (Voelkel 1993; Stehr 1996). The concept was based on the UV-
absorption of the carboxyl-residues of the amino acids at 210 nm. A major problem was
the absorption of other media components in this wavelength range. Evidence had shown
that this was no problem for biotransformation analysis but for analysis of fermentation
and cultivation samples because of the manifold composition of some complex media. An
advantage of this method was the fast sample preparation. The samples were first centri-
fuged (15 min, 12,000 rpm; Heraeus Christ Biofuge A), then the supernatant was measured
in the HPLC system. It was possible to measure the concentrations of the product L-
methionine, the substrate DL-MTEH and the intermediate D-carbamoyl-methionine (see
Figure 2.4). The components and the optimised conditions are shown in Table 3.8. The

- 41 -
_________________________________________________________________________
calculation of the concentrations is based on calibration curves; solutions in concentration
ranges between 0.1 g/l and 1 g/l were used.

Table 3.8: Components of the HPLC system and conditions for the optimised separation.

process performance isocratic, room temperature
sample volume 20 l
mobile phase 1% acetonitrile, 1% trifluoroacetic acid;
both dissolved in H
2
O, pH 2
flow rate 0.8 ml/min
column
(stationary phase)
RP-18 column (Nucleosil), 250 mm length,
4mm diameter
components
HPLC system
UV detector (model SPD-6AV, Shimadzu)
integrator and control system (Andromeda 1.6)
automated sampler (model 816. Knauer)
pump (model 880 PU, Jasco)

3.4.6 Gas chromatography (GC)

The gas chromatography (GC) system (type Hewlett Packard HP5890 with automated
sampler) is the most important analysis method for quantitative analysis of amino acids.
The sensitivity is very high at concentration ranges above 1 mg/l. A disadvantage was the
long time of sample preparation due to the derivatization process which was necessary for
the measurement with this GC method. The samples were centrifuged (15 min, 12,000
rpm; Heraeus Christ Biofuge A), then the supernatant was used for the derivatization proc-
ess with a special kit from the Phenomenex company (EZ:faast for free physiological
amino acid analysis by GC-FID (gas chromatography flame ionization detector)). The in-
tention of the derivatization was the transfer of amino acids from the aqueous sample to
special organic solvents (in this case a mixture of octane and chloroform, the mobile
phase) and the modification of the amino acids with special chemical groups in order to
increase their solubilities in the organic mobile phase and to make them volatile, which
was a requirement for measuring in GC systems. The derivatization kit contained an inter-

- 42 -
_________________________________________________________________________
nal standard, norvaline, in a concentration of 200 mM. This is an important factor for the
interpretation of results, because the concentration of other substances were determined in
comparison to the known concentration of the internal standard. Details about the process
are shown in Table 3.9. Another important point for the process was the temperature pro-
gram for the heater (oven) to get stable base lines.

Table 3.9: Components of the GC system and conditions for the process performance.

process performance
(gas flow, mobile phase)
30 ml/min N
2
, 30 ml/min H
2
,
300 ml/min synthetic air
analysed sample volume 1-2 l
further adjustments detector temperature 280C,
injector temperature 250C
temperature program
(oven)
110 300C with 32C/ min
column
(stationary phase)
Chrompack CP-Sil-8CB, low bleed capillary
column (15 m 0.25 mm 0.25 m)
detector type FID (flame ionization detector)
components GC system GC system type Hewlett Packard 5890A
front injector (HP7673A), automated sampler
computer control system: GC ChemStation
Rev.A.10.02 [1757], Agilent Technologies

The analysis kit needs to be stored at 4C, it consists of 6 chemical solutions. The solutions
and the ingredients are described in Table 3.10. In general, it is always important to mix
the liquid substances for 5 8 seconds and to wait for at least one minute (diffusion proc-
ess) before proceeding with the instruction manual (Phenomenex 2005).

- 43 -
_________________________________________________________________________
Table 3.10: Reagent composition for derivatization (Phenomenex, 2005)

Reagent Ingredients
1: internal standard solution norvaline (0.2 mM), N-propanol 10%
2: washing solution N-propanol
3A: eluting medium component 1 sodium hydroxide
3B: eluting medium component 2 N-propanol
4: organic solution 1 chloroform
5: organic solution 2 iso-octane
6: re-dissolution solvent iso-octane 80%, chloroform 20%

The first step of the derivatization was the production of fresh eluting medium 3 by pool-
ing the reagents 3A and 3B (see Table 3.10) in a ratio of 3:2. A 100 l sample was pooled
with 100 l reagent 1 in a sample preparation vial and mixed. This step guaranteed a pH
around 7 and also the internal standard norvaline was given to the sample. In the second
step the solution in the sample preparation vial was passed through a sorbent tip using a
syringe. The tip contained solid material which was capable of binding the amino acids. In
the next step 200 l of reagent 2 were passed through the sorbent tip (syringe). The liquid
in the syringe (reagent 1 + 2) was discarded, then 200 l of reagent 3 was given in the
sample preparation vial and passed through the sorbent tip (syringe) until the liquid
reached the filter plug in the sorbent tip. Then both the liquid (reagent 3) and the sorbent
particles were ejected out of the tip into the sample preparation vial. The next step was the
consecutive addition of reagent 4 (50 l) and reagent 5 (100 l) with a special pipette de-
livered with the kit (Drummond Dialamatic Microdispenser) to avoid cross contamination.
Finally 100 l of reagent 6 was added to the sample preparation vial. The reagents 4, 5 and
6 contained organic solvents (chloroform + iso-octane, see Table 3.10). These substances
were added to extract the formed amino acid derivatives from the aqueous phase to the
organic phase. The organic phase (upper phase) was transferred to an autosampler vial for
the measurement with the GC system (for further and detailed information about the deri-
vatization, see the instruction manual for the analysis kit from the Phenomenex company).
An amount of 2 l of each sample was injected and measured with the gas chromatography
system.

- 44 -
_________________________________________________________________________

reagents 1-6
micro-
dispenser
syringes
sorbent tips
sample
preparation
vial
Figure 3.5: View on the EZ:faast kit for free physiological amino acid analysis by GC-FID (Phe-
nomenex company).

3.4.6.1 Gas chromatography - mass spectrometry (GC-MS)

The combination of gas chromatography (GC) and mass spectrometry (MS) is used to ana-
lyse mixtures. The GC serves to separate the sample components, the mass spectrometry
serves to identify the constituents in context with the retention data. Principle of the MS is
in general an electronic ionisation of the chemical substances and the formation of charac-
teristical fragments which can be detected and identified on a photo plate.

The GC-MS experiments for this research were performed at the Institute for Organic
Chemistry, TU Braunschweig by Dr. Till Beuerle. The used gas chromatography was an
Agilent 6890 gas chromatograph, equipped with a 30 m analytical column (Phenomenex
ZB5-MS, 30 m 0,25 mm ID). The electronic ionisation (EI) mass spectrometer was a
JMS-T100GC (GCAccuTOF, JEOL, Japan) model.

3.5 Bioreactor experiments

The bioreactor experiments were performed in 3 and 3.5 L scale in glass reactors (com-
pany: Infors, type: Minifors, reactor volume 5 L). A single-wall glass vessel (see Figure
- 45 -
_________________________________________________________________________
3.6) with a stainless steel lid served as bioreactor. The electrodes [pH, temperature, anti-
foam and pO
2
(oxygen partial pressure)], the sampling system, the stirring shaft and the
connection to the stirring motor, the equipment for aeration, cooling, exhaust air and inlets
for the supply of the reactor content with acids, bases, anti-foam and feeding substances
over pumping systems were fixed in the lid. The automatic control technique was able to
measure and to control temperature, stirring speed, pH, pO
2
(oxygen partial pressure) and
foam development. The pO
2
value could be combined with the stirring speed (stirrer cas-
cade): If the value of the oxygen parial pressure dropped under 20%, the stirring speed
would automatically increase to improve the oxygen supply in the bioreactor. The stirrer
was a 6 blade disk impeller type (Rushton turbine).

rotameter aeration control
Stainless steel lid with equipment
(electrodes, inlets, connection stirring
motor to stirring shaft, cooling sys-
tem, aeration, exhaust air, sampling)
automatic control technique
with display unit
heating / cooling plate connected with
heat and cooling system
single-wall glass vessel
sampling bottle
bottles for addition / feeding with pump
system (acid, base, glucose, anti-foam)

Figure 3.6: Picture of the bioreactor (company Infors, type Minifors) and the automatic control
unit (controlling of temperature and cooling system, aeration rate, foam sensor, pumping systems
for acids, bases, anti-foam substances and fed batch supply).

The online data acquisition was made by the computer program IRIS VIEW, the program
was able to aquire the data for pO
2
, pH, stirring speed, temperature and pumping rates of
- 46 -
_________________________________________________________________________
feeding (acid, base, anti-foam substance). The dimensions of the bioreactor system are
shown in Figure 3.7 and the following Table 3.11.

D
d
d1
h
2
H
h
h
2
h
1
Gassing

Figure 3.7 + Table 3.11: Schematic presentation of the bioreactor dimensions and register of the
corresponding values.

Description Abbreviation Numerical value (cm) /
number
Reactor fill level H 25 cm (3.5 L)
Total reactor volume h 33 cm (5 L)
Working volume V 3 3.5 L
Reactor diameter D 13.5 cm
Diameter of stirrer d 6 cm
Diameter stirring shaft d1 1.2 cm
H/D ratio H / D 1.85
6-Blade disk impeller - 2
Stirrer fitting level 1 h1 8.5 cm
Stirrer fitting level 2 h2 15.5 cm
Gassing vents - 5

- 47 -
_________________________________________________________________________
In general the cultivation parameter for the fermentation with Corynebacterium glu-
tamicum strains (KY10574 + DSM20300) were aquired and controlled in the following
ways: temperature (30C), pH (only aquisition), stirring speed (700 1,000 rpm), aeration
rate: 0.5 L/(L min), automatic increased stirrer speed by oxygen partial pressure under
20% (stirrer cascade), anti-foam monitoring and automatic addition of anti foam substance
Struktol J673 (Schill + Seilacher company; dilution with water 1:10).

3.5.1 Exhaust gas analysis

The analysis of exhaust gases was measured by 2 sensors which were connected with the
computer program FERMVIS. The oxygen uptake rate (OUR) and the carbon dioxide evo-
lution rate (CER) can be counted by the method of the quasi-stationary inert gas calcula-
tion if the concentration of oxygen and CO
2
in fresh and exhaust air are known. Addition-
ally, the aeration rate and the liquid volume in the bioreactor have to be known (Chmiel
2006). The respiratory coefficent RQ is calculated as the coefficient q
co2
/ q
o2
(referring to
molar mass).

( )
( )

=
A
CO
A
O
E
O
E
CO
E
O
A
O
N
F
O
E
O
N N
O
X X X
X X X
T R V
MG X P V
Q
2 2 2
2 2 2 2 2
2
1
1
1
&

( )
( )

= 1
1
1
2 2 2
2 2 2 2 2
2 A
CO
A
O
E
CO
E
CO
E
O
A
CO
N
F
CO
E
CO
N N
CO
X X X
X X X
T R V
MG X P V
Q
&

Used abbreviations und indices:

V [l h
-1
]
V
F
[l]
P [bar]
R = 8.314 10
-2
bar l K
-1
mol
-1

T [K]
general gas constant
temperature
MG [g mol
-1
]
X
E
X
A

N
flow rate
volume of liquid in the bioreactor
pressure
molar mass
mole fraction in fresh air (inlet)
mole fraction in exhaust air (outlet)
standard state (273.15 K, 1 bar)

- 48 -
___________________________________________________________________________
3.6 Screening and selection system

The screening for L-methionine producing mutants was performed on the basis of Corynebac-
terium glutamicum DSM20300. The preparation of microorganisms in order to radiate them
with UV-light was performed in the following steps (Gerhardt et al. 1981):

First, 50 ml SMM medium containing the strain Corynebacterium glutamicum DSM20300
was cultivated to a cell density of approximately 2 10
8
cells / ml. Then the culture was
cooled on ice (10 minutes) to prevent further growth. Afterwards the culture was divided in 5
ml portions and centrifuged for 5 minutes at 6,000 rpm (Heraeus Sepatech Labofuge A) to
pellet the cells. Each cell pellet was resuspended in 2.5 ml of sterile MgSO
4
solution (0.1 mo-
lar). The cultures were transferred to sterile Petri plates (100 mm diameter). The plates were
placed on a compensator in order to move the liquid content slightly. The compensator was
set up under a specially reconstructed hood (see Figure 3.8) with an UV lamp (wavelength
254 nm).

UV-radiation of liquid cultures on a
shaker (max. 2 min)
agar plate:
minimal medium
+ antimetabolite
(ethionine x g/l)
cultivation of mutant colonies
in minimal media (150 l/well),
(24 h, 1200 rpm, 30C)
transfer to
agar minimal media
s se el le ec ct ti io on n s st te ep p
transfer in 96-well-plates

centrifugation

amino acid
analysis
TLC-ninhydrin reaction (color reaction + heat)

Figure 3.8: Schematical course of action for the screening experiments in order to obtain L-
methionine overproducing strains (UV-radiation, selection, cultivation, analysis).

- 49 -
___________________________________________________________________________
UV-radiation was optimised to a bacterial killing rate of approximately 99.9% in order to en-
hance the probability of producing bacterial mutant which had the metabolic modifications for
L-methionine overproduction. The distance between the height of the rocker and the height of
the UV-lamp were 21 cm; under these conditions the survival rate of Corynebacterium glu-
tamicum DSM20300 dropped under 0.1% after 55 seconds (see Figure 3.9).

0,01
0,1
1
10
100
0 20 40 60 80 100 12
Time of radiation [s]
b
a
c
t
e
r
i
a
l

s
u
r
v
i
v
a
l

r
a
t
e

[
%
]
0

Figure 3.9: Survival rate [%] of Corynebacterium glutamicum by UV-radiation (254 nm) in depend-
ence of the radiation time

After random mutation with UV-radiation the liquid was transferred to agar plates with SMM
minmal medium plus different concentrations of anti-metabolite ethionine (in the most cases
2.5 g/l) in order to select a large amount of obtained bacterial mutants. The agar plates were
incubated at 30C for 48 h. It had been expected that only ethionine resistant mutants would
be able to grow on these agar medias (see chapter 2.5) and, additionally, that these mutants
would be capable of overproducing L-methionine. The growing colonies on the agar plates
were separated by transferring the material with inoculating-loops in 96-well plates with 150
l/well of SMM medium. The 96-well-plates were incubated on a shaker (see Figure 3.8) for
24 h by 30C with 1200 rpm. The incubated samples were centrifuged with 6,000 rpm (Ep-
pendorff Centrifuge 58R). In a final step, an aliquot of 5 l was transferred under sterile con-
ditions on thin layer chromatograpy (TLC) aluminium sheets for fast amino acid analysis. The
TLC analysis was used because of the large number of samples which made it necessary to

- 50 -
___________________________________________________________________________
measure in a fast and efficient way (detailed information about the thin layer chromatography
(TLC) see chapter 3.4.4).
3.7 Cell disruption of C. glutamicum strains

The cell disruption with Corynebacterium glutamicum strains (KY10574 + DSM20300) was
performed to measure the intracellular enrichment and concentration of L-methionine. The
biomass was taken from bioreactor experiments after 48 h to get enough comparable material
for several reproducible experiments.

As a first step the cultivation broth (Sorvall RC 5B plus centrifuge) was centrifuged for 20
min. with 12,000 rpm to pellet the biomass. The biomass was washed with sterile distilled
water and centrifuged again under the same conditions. Then the biomass was weighed and
the same amount of distilled water was added in order to get cell suspension. The cell suspen-
sion was disrupted by an ultrasonic homogenizer (Bandelin sonoplus HD2200 ultrasonic
sonotrode type MS 72 and MS73), the disruption time was about 15-20 minutes. Every other
minute a sample was taken for Bradford analysis to provide data of the disruption level (see
chapter 3.7.1).

In a second step, the water was removed from the sample by lyophilisation. This was per-
formed by a freeze drying system of the Omnilab company (type Alpha 1-4 LDplus) for 24 h.
The obtained bio dry mass was weighed and resolved in a small amount of distilled water in
order to concentrate the intracellular substances (especially the amino acids). The gained sus-
pension was ultrafiltrated in order to cut off the higher molecular weights (MWCO). This fil-
tration was performed by centrifugation (5,000 rpm, centrifuge Heraeus Sepatech Labofuge
A) with a filtration vessel (Vivascience company, type Vivaspin 6. MWCO: 50 kDa). The
cleaned solution was measured with gas chromatography (GC).

For Corynebacterium glutamicum DSM20300 the cell disruption was also performed using a
swing mill (Retsch company, type MM2) with metal beads as grinding bodies (0.75 mm di-
ameter). The process was carried out for 20 minutes with a grinding body filling degree of
52.5%. The cell suspension was made in the same manner as described earlier for the ultra-
sonic disruption.

- 51 -
___________________________________________________________________________
3.7.1 Bradford assay

The protein measurement according to the Bradford method was an important step in order to
determine the degree of cell disruption. During the disruption process the proteins were re-
leased from the cytosol and could be detected by this analysis. Therefore the protein concen-
tration increased until the cell disruption degree reached nearly 100%.

The measurements were carried out with Bradford solution (Bio-Rad protein assay solution).
1,000 l of Bradford solution (diluted in a 1:5 ratio with distilled water) were mixed with a 20
l sample. After 20 minutes the sample was measured photometrically (Photometer, Phar-
macia Biotech company, type Ultrospec 3000) in a 1 ml cuvette at 595 nm. The calibration
was performed with bovine serum albumin (BSA) solutions in concentrations ranges from 0.1
to 0.6 g/l (linear area).

3.8 Protein hydrolysis of C. glutamicum KY10574

The protein hydrolysis of freeze-dried biomass from Corynebacterium glutamicum KY10574
and the following determination of amino acids (methionine, lysine and threonine) was per-
formed at the LUFA (Landwirtschaftliche Forschungs- und Untersuchungsanstalt) in Speyer.

3.9 Chemicals and equipment

The chemicals described in chapter 3 were all purchased from the following companies in
Germany: Sigma Aldrich, Fluka, Merck, Roth. The L-methionine analogue ethionine was
bought from the ABCR company. The anti-foam reagent Struktol J673 was purchased from
the Schill and Seilacher company. The used equipment was directly named in the context of
its application. The substrates (MTEH) and the inducer (IMH) for the biotransformation ex-
periments were received from the Evonik (former Degussa) company.

- 52 -
Results and discussion
___________________________________________________________________________
4 Results and discussion

4.1 Arthrobacter aurescens DSM7330: L-methionine pro-
duction with biotransformation

In order to become familiar with L-methionine production and analytics a well known bio-
transformation was performed (Voelkel 1993; Stehr 1996). Using Arthrobacter aurescens
DSM7330 DL-MTEH was converted into L-methionine (see Figure 4.1). Since DL-MTEH
was derived chemically this procedure is not in agreement with the side conditions of our pro-
ject (new European law; see introduction chapter). The analytics were performed with HPLC.
Time (h)
0,0 2,5 5,0 7,5 10,0 12,5 15,0 17,5 20,0 22,5 25,0
L
-
m
e
t
h
i
o
n
i
n
e

(
m
g
/
l
)
0
50
100
150
200
250
300
350

Figure 4.1: Biotransformation of the substrate DL-MTEH to the product L-methionine with whole
cells of Arthrobacter aurescens DSM7730; biocatalysis was performed in 100 ml shake flasks (9 ml
substrate solution, 1 ml of 50% cell suspension, 37C, 24 h, 100 rpm); quantitative analysis with
HPLC; for detailed information see chapters 3.2 and 3.4.5.

Figure 4.1 shows that the highest value of L-methionine production was reached after 5 h with
324 mg/l. This corresponded with a yield of 0.35 g/l L-methionine / g DL-MTEH. In former
studies (Stehr 1996) it was proved that even higher yield values were possible after several
optimisation experiments. After 24 h there was no further production; on the contrary, the L-

- 53 -
___________________________________________________________________________
methionine concentration dropped below 300 mg/l, the balance between substrate and product
became more and more inappropriate.

Although the used HPLC method was very well established for the analyzation of this bio-
transformation experiment, it turned to be out in the following time that this method is not
suitable for quantitative amino acid measurement of biological cultivation. The reason is the
manifold absorption appearance of medium components and byproducts at the measurement
wavelength of 210 nm. For this reason in the following experiments with Corynebacterium
glutamicum the gas chromatography method (see chapter 3.4.6) was developed and used.

The enantioselectivity of the biotransformation reaction is shown in Figure 4.2 with a chiral
thin layer chromatography (TLC). By comparison of the references for D + L-methionine
Figure 4.2 points out clearly that the enzymes from Arthrobacter aurescens DSM7330 cells
were able to convert the L-enantiomeric substrate form to L-methionine whereas D-
methionine was not produced.

L-methionine
0.5 g/l
D-methionine
0.5 g/l
Sample biotransformation
5h, 324 mg/l L-methionine

Figure 4.2: Qualitative chiral thin layer chromatography (TLC); the sample after 5 h of biotransfor-
mation (see Figure 4.1) and two reference solutions for D- and L-methionine (0.5 g/l) are illustrated,
detection by ninhydrin reaction.

- 54 -
___________________________________________________________________________
4.2 Corynebacterium glutamicum DSM20300

4.2.1 L-glutamic acid production with C. glutamicum DSM20300

The strain C. glutamicum DSM20300 is well-known as a producer of L-glutamic acid (Kino-
shita et al. 1957). In this thesis the strain was important for the screening on L-methionine
overproducers. Therefore it was obvious to test the quality of the strain material in relation
with its most obvious characteristic feature, the L-glutamic acid production capability, by
trying to reproduce former results from literature. The experiments were carried out in the
shake flask scale with a complex preculture and a minimal main medium (see chapter 3.3.2
Table 3.3); the results are shown in Figure 4.3.
Time (h)
0 20 40 60 80 100 120 140 160
B
D
M

(
g
/
l
)
0
1
2
3
4
5
6
7
G
l
u
c
o
s
e

(
g
/
l
)
25
30
35
40
45
50
55
L
-
G
l
u
t
a
m
i
c

a
c
i
d

(
g
/
l
)
0,0
0,2
0,4
0,6
0,8
1,0
1,2
BDM
Glucose
L-Glutamic acid

Figure 4.3: Cultivation of Corynebacterium glutamicum DSM20300 in the shake flask scale in L-
glutamic acid production medium (500 ml shake flask, 2 baffles, 30C, 100 rpm, 100 ml volume of
MPM medium; inoculated with 2 ml from a 17 h old preculture with complex ingredients (CP)). The
Figure shows the curve progressions of glucose, BDM and glutamic acid values in main medium
MPM; amino acid analytics with GC.

It can be seen that the production of L-glutamic acid in the main medium increased rapidly in
the first hours of cultivation time, especially around 20 hours. The production of L-glutamic
acid began nearly without delay due to the use of the complex preculture (CP medium, Table

- 55 -
___________________________________________________________________________
3.3). In the same manner a high formation of bio dry mass was observed in the first 20 hours.
After 50 hours the stationary phase was achieved and the bio dry mass stayed constant at 6
g/l. The glucose was not consumed any more. The L-glutamic acid concentration was around
900 mg/l after 70 hours, the concentration increased slightly to 1 g/l in the following 50 hours,
perhaps due to the L-glutamic acid from the intracellular milieu. Probably the contents from
this milieu were released because of the deficiency of nutrients (carbon, sulphur, nitrogen and
ammonia). An enrichment of intracellular L-glutamic acid in Corynebacterium glutamicum
was observed in former investigations (Clement et al. 1986). The glucose level after 70 hours
remained constant, so the nutrient deficiency has to be located in the other sources mentioned
above.

In the literature (Kinoshita et al. 1961) L-glutamic acid concentrations around 3 g/l were ob-
tained using this strain. The first basic approach to improve this experiment has to be the pre-
vention of the nutrient deficiency after 70 hours. In further experiments higher concentrations
of L-glutamic acid can be also received by optimising the cultivation parameter.

4.2.2 L-methionine production capability of C. glutamicum
DSM20300

The wild type Corynebacterium glutamicum DSM20300 is not able to produce L-methionine
in higher amounts than necessary for its own survival. The feedback inhibition and repression
signalling cascades are fully activated (see Figure 2.10), they prevent the unefficient enrich-
ment of metabolites which are not directly useful for the bacterial growth and survival. As
preliminary investigation, before the beginning of the screening experiments, it was important
to get results about the natural L-methionine production abilities.

In Figure 4.4 three representative cultivations with Corynebacterium glutamicum DSM20300
are illustrated. All cultivations were performed in shake flasks with the same conditions, the
used modified SMM medium (see Table 3.2) was also used in the screening experiments (see
next chapter). The direct metabolic precursors of L-methionine, L-homoserine and L-
homocystine, were put to the SMM medium in order to support the L-methionine production
and to diminish the feedback inhibition and repression phenomenons (see Figure 2.10).

- 56 -
___________________________________________________________________________

0
2
4
6
8
10
12
24 48 72
Time (h)
L
-
M
e
t
h
i
o
n
i
n
e

(
m
g
/
l
)
without precursor
+0.5 g/l Homoserine
+0.5 g/l Homocystine

Figure 4.4: Cultivation of Corynebacterium glutamicum DSM20300 in the shake flask scale in SMM
medium (100 ml SMM medium in 1 L shake flask (2 baffles), 30C, 120 rpm; inoculated directly from
BY agar plates; the L-methionine precursors in metabolic pathway, homoserine and homocystine were
given to the SMM medium; amino acid analysis with GC.

In pure SMM medium 5.5 mg/l L-methionine was measured after 24 h, which decreased in
the next 48 h below values of 4 mg/l. In the medium with 0.5 g/l homoserine 11 mg/l L-
methionine were produced in the first 24 h, after that L-methionine was metabolized to 4 mg/l
in the next 48 h. The results for the medium with the addition of 0.5 g/l homocystine were
comparable to the results with the pure SMM medium. Therefore, it can be concluded that the
addition of metabolic precursors of L-methionine does not result in a notable enhancement of
L-methionine production. Only in the first 24 h a slight accumulation of L-methionine in the
cultivation with the SMM medium plus homoserine was observed, but this product was con-
sumed in the following cultivation time. To summarize, it can be pointed out that the addition
of precursors was not appropriate to increase the production level of L-methionine signifi-
cantly. The basis L-methionine level around 4 mg/l is necessary for the microorganism to
provide his own metabolism with the essentiell amino acid L-methionine.

- 57 -
___________________________________________________________________________
4.2.3 Screening for L-methionine overproducers with C. glutamicum
DSM20300 with random mutagenesis and following selection

The main problem of the screening system was the selection process after UV-radiation. The
UV-radiation was working well and quickly (see bacterial killing curve in Figure 3.9). The
high amount of obtained mutants was in fact the real problem. The selection system is based
on the theory that only mutants which are resistant to the L-methionine analogues (e.g. ethion-
ine) are able to grow on media with these analogues and that these mutants have the capability
of producing L-methionine.

The colony growth of Corynebacterium glutamicum DSM20300 is shown in Figure 4.5 in
comparison to the concentration of L-ethionine. The influence of L-ethionine concentration
on colony growth inhibition began at a concentration of 2 g/l. A slight reduction of the colony
amount was observed in the first 24 h (85%), but the amount of colonies increased again on
the following days to approximately 90%. There was a reduction of colony growth to 50%
using 5 g/l of ethionine in the first 24 h, but only the high concentration of 10 g/l ethionine
could prevent the colony growth in the first 24 h totally. On the second day the amount of
colonies increased again over values of 50%. L-ethionine is a very expensive substance, the
use of 10 g/l cannot be recommend under economic considerations for laboratory experi-
ments. In addition, concentrations around 10 g/l L-ethionine did not only have effects on L-
methionine but poisonous impacts on the whole biosynthesis. There were also attempts to
work with other anti-metabolites (methyl methionine, norleucine), but the potency of these
substances in growth inhibition was even lower. In the screening system L-ethionine concen-
trations of 2.5 g/l in SMM medium were used as selection step after UV-radiation as a com-
promise between economic costs and inhibition effects.

- 58 -
___________________________________________________________________________
0
10
20
30
40
50
60
70
80
90
100
0 1 2 3 4 5 10
Ethionine (g/l)
c
o
l
o
n
y

g
r
o
w
t
h

(
%
)
colonies (%), 24 h
colonies (%), 48 h
colonies (%), 72 h

Figure 4.5: Cultivation of Corynebacterium glutamicum DSM20300 on agar plates in SMM agar me-
dium with different concentrations of the L-methionine anti-metabolite / structural analogue L-
ethionine; presented is the dependency of colony growth inhibition after 24 h, 48 h and 72 h with the
concentration of L-ethionine.

The first experiments resulted in another problem. The screening was performed as described
in chapter 3.6. The assumption that all colonies which are able to grow on L-ethionine are
also L-methionine overproducers proved to be wrong. The theoretical mechanism which leads
to this theory is described in chapter 2.5. The resistance of the colonies to L-ethionine is a
necessary criterion, but - and this is the crucial point - not a sufficient criterion. It is therefore
unevitable to measure the whole amount of colonies which are able to grow on media with
distinct concentrations of L-ethionine. It was attempted to optimise the efficiency of this proc-
ess and a measurement throughput of 500 600 samples per week was reached. But this was
not enough to get a real chance to find an overproducer due to the millions of colonies re-
ceived after the UV-radiation. An attempt to pick up the first appearing colonies on the agar
plates for the first measurements with thin layer chromatography (TLC) assuming that they
were most of all capable of producing L-methionine was not successful.

A typical screening result obtained with thin layer chromatography (TLC) is shown in Figure
4.6. The spot on the left side (L-methionine 1 g/l) and the other one on the right side (L-
glutamic acid 1 g/l) of the TLC are standards, the other samples in the middle represent strain
material of C. glutamicum DSM20300. This material was treated with UV-radiation and fol-
lowing selection by cultivating in medium in the presence of the anti-metabolite L-ethionine.
The TLC is a fast visual analysis method. Further precise analysis with gas chromatography

- 59 -
___________________________________________________________________________
might be used if the results in TLC indicated L-methionine production. The comparison of the
spots in Figure 4.6 shows that there is production of L-glutamic acid, but no production of L-
methionine. The wild type strain DSM20300 is able to produce L-glutamic acid, thus it is
possible and probable that the derived mutants keep this characteristic feature after UV-
radiation treatment.

C. glutamicum DSM20300 strain mate-
rial after UV-treatment and selection
with ethionine
L-glutamic
acid, 1 g/l
L-methionine
1 g/l

Figure 4.6: Qualitative thin layer chromatography of Corynebacterium glutamicum DSM20300 strain
material samples after treatment with UV-radiation and anti-metabolite treatment (ethionine); stan-
dards are solutions of L-methionine and L-glutamic acid (1 g/l), detection with ninhydrin.

There might be better chances to develop a L-methionine overproducer with the strain Cory-
nebacterium glutamicum KY10574. This strain is able to produce L-methionine in higher
amounts (see following chapters). It could be easier to improve the properties of this strain by
random mutation methods than to develop an overproducer from strain material which is not
able to produce L-methionine in notable amounts. The feedback inhibition and repression
signal system in C. glutamicum KY10574 has been already deregulated, for this reason the
chances to get further success in strain treatment with random mutation increase.

- 60 -
___________________________________________________________________________
4.3 Corynebacterium glutamicum ATCC21608

There were several experiments in the shake flask scale to observe the amino acid production
of the potential L-methionine producer Corynebacterium glutamicum ATCC21608. In the
following shake flask cultivations of this strain in F1 medium (see Table 3.6) and in KH1 /
KH2 medium (see Table 3.4 / 3.5) are presented and explained. A growth of this strain in pure
SMM minimal medium could not be observed, the strain demanded for more nutrients as
available in basic mineral media. This is a typical side effect of random mutation procedures,
the probability of additional unrequested mutations is extremely high. As a result the new
strain shows in many cases growth deficiency in basic minimal media. This was not a major
problem in the cultivation with the KH1 / KH2 media because the KH1 medium for the pre-
culture had complex ingredients and the KH2 medium was inoculated in the next step with
KH1 complex medium. The F1 medium is a minimal medium, but it contains many additional
ingredients like mineral elements, vitamins and protocatechic acid, a chelating compound
which can be very useful for enhanced growth of Corynebacteria cell mass (Liebl 1989).

The cultivation of the strain Corynebacterium glutamicum ATCC21608 in a complex precul-
ture KH1 and following KH2 minimal medium is shown in the Figure 4.7. The minimal me-
dium was inoculated with the preculture in a ratio of 1:11. Therefore, the curves do not start at
zero in the graphs. The preculture contained particularly L-lysine in higher amounts around
300 mg/l after 24 h (no Figure), but there was no L-lysine production during the following
cultivation in minimal medium anymore (see Figure 4.7).

The impact of amino acid components of the complex media on the process of production is
always difficult to predict and hardly describable. It is also possible that there are unrepro-
ducible conversions of complex media components during the cultivation time which can in-
crease the concentration of amino acids temporarily. Thus it is complicated to optimise a
complex medium for microbial overproducers of amino acids due to nonexistent knowledge
about the exact composition of every new media charge. Additionally, the costs for industrial
process performance are much higher using complex ingredients.

- 61 -
___________________________________________________________________________
Time (h)
0 20 40 60 80 100 120 140
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
25
50
75
100
125
150
175
200
225
Alanine
Methionine
Lysine
Glycine

Time (h)
0 50 100 150 200 250
B
D
M

(
g
/
l
)
0
2
4
6
8
10
12
G
l
u
c
o
s
e

(
g
/
l
)
0
20
40
60
80
100
p
H
5,0
5,5
6,0
6,5
7,0
7,5
8,0
8,5
9,0
BDM
Glucose
pH

Figure 4.7: Amino acid production and development of bio dry mass (BDM), glucose and pH during
the cultivation of C. glutamicum ATCC21608 in KH2 medium (100 g/l glucose) over 144 h in the
shake flask scale (110 ml in 1L volume shake flask, 2 baffles, 120 rpm, 30C); inoculated with 10 ml
from complex preculture KH1 (100 ml in 1 L volume shake flask, 2 baffles, 120 rpm, 30C, 24 h);
amino acid analytics with GC.

- 62 -
___________________________________________________________________________
Production of L-methionine could not be observed in levels higher than necessary for the sup-
port of the own metabolism of the strain C.glutamicum ATCC21608. On the other side the
amino acid L-alanine was produced in concentrations around 200 mg/l after 96 h. This was
also observed in another minimal medium (F1), shown in Figure 4.8.

The connection between L-alanine production and the glucose amount is quite clear to see in
the Figures 4.7 and 4.8. L-alanine was produced as long as the glucose was not consumed
totally. After that the L-alanine concentration began to drop quickly whereas the glycine con-
centration began to increase. This phenomenon also took place with another medium (F1), to
be seen in Figure 4.8. But in that case there was also a high increase of L-lysine concentration
at the end of the cultivation.

It could be possible that L-alanine is converted to glycine by cutting off the methyl group in
order to supply the strain with carbon due to the background that glucose as the original car-
bon source has been consumed totally at this point in time. Under conditions of nutrient defi-
ciency amino acids are degraded and converted in order to maintain the metabolic activities.

L-alanine and glycine are glucogenic amino acids, they are degraded through transamination
to pyruvate (see Figure 2.12). Pyruvate can be converted in the next step by gluconeogenesis
to glucose. A conversion from alanine to glycine is supported by these facts according to the
same biochemical degradation pathway.

In contrast, L-lysine is a ketogenic amino acid (Figure 2.12), the carbon atoms are degraded to
acetyl-CoA and acetoacetate. Humans and animals do not possess a metabolic pathway for the
conversion of these biochemical products to glucose (Voet et al. 2002). Therefore, a conver-
sion from L-alanine to L-lysine under glucose limited conditions is very improbable.

The pH graph in Figure 4.7 showed a characteristic course, it decreased during glucose con-
sumption, probably according to the output of products with slightly acidic properties. Then,
the pH increased rapidly in the basic area around 8.5. There are several reasons for this like
the release from basic cell contents after cell collapsing and the production change to other
substances under carbon source limited conditions.

- 63 -
___________________________________________________________________________
The stagnation of the bio dry mass as shown in Figure 4.7 (stationary phase began at 75 h)
was nearly parallel to the complete glucose consumption. The cells were not able to prolifer-
ate under carbon limiting conditions. The fast bio dry mass development during the first 96 h
showed the growth ability of these microorganisms in the exponentiell growth phase using the
available nutrients in an efficient way. The measurement of the BDM, however, turned out to
be complicated due to the unsoluble CaCO
3
. This led to inaccurate results in BDM, and thus
the measurement of the optical density (OD) was not performed.

In further experiments it could be observed that the strain Corynebacterium glutamicum
ATCC21608 was able to grow on F1 minimal medium which was originally designed to cul-
tivate the strain Corynebacterium glutamicum KY10574 (see later). The results are shown in
Figure 4.8. The strain ATCC21608 showed similar production properties in this medium
compared to the cultivation in KH2 medium. L-alanine was produced in concentrations
around 175 mg/l at the top after 72 h as long as the carbon source glucose was not consumed
totally. Afterwards the L-alanine concentration began to drop, on the other hand the L-lysine
concentration began to rise.

Additionally, a remarkable point is the cell growth velocity in the F1 medium (see Figure
4.8). In comparison to the cultivation shown in Figure 4.7 higher bio dry mass values could be
reached in this experiment (13 g/l in maximum after 50 h compared to 10.5 g/l after 75 h in
the former shake flask cultivation). The cultivation in F1 minimal medium outstripped the
combination of complex preculture (KH1 medium) and minimal production medium (KH2
medium) in bio dry mass formation and cell growth velocity. Therefore it can be concluded
that the F1 minimal medium contents are very well adapted to the nutrient demands of C. glu-
tamicum ATCC21608 cells.

Higher L-alanine concentrations might be possible with more glucose in the F1 minimal me-
dium (20 g/l in this case, Figure 4.8). But there were no higher results in L-alanine production
(around 200 mg/l) using KH2 medium (Figure 4.7), which contained 100 g/l glucose. There-
fore, it would be more promising to investigate other basic nutrient concentrations in the me-
dia like phosphate, sulphate or nitrogen in order to enhance the L-alanine production with
Corynebacterium glutamicum ATCC21608.

- 64 -
___________________________________________________________________________
Time (h)
0 20 40 60 80 100
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
20
40
60
80
100
120
140
160
180
200
Methionine
Alanine
Lysine
Glycine

Time (h)
0 20 40 60 80 100
G
l
u
c
o
s
e

(
g
/
l
)
0
5
10
15
20
25
B
D
M

(
g
/
l
)
0
2
4
6
8
10
12
14
Glucose
BDM

Figure 4.8: Amino acid production and development of bio dry mass (BDM) and glucose during the
cultivation of Corynebacterium glutamicum ATCC21608 in the shake flask scale in F1 medium with
20 g/l glucose (100 ml F1 medium in 1 L volume shake flask, 2 baffles, 30C, 120 rpm); directly in-
oculated with a loopful of cells from BY agar plates; amino acid analysis with GC.

- 65 -
___________________________________________________________________________
On an industrial scale L-alanine is produced from DL-aspartic acid by means of immobilized
Pseudomonas dacunhae cells, the responsible enzyme is the L-asp--decarboxylase (Furui et
al. 1983). Most organisms are able to produce DL-alanine in direct fermentation due to the
existence of the enzyme alanine racemase. In 1993 a report was published describing the pro-
cedure to get 46 g/l D-alanine from Bacillus lactofermentum (Yahata et al. 1993). L-alanine
can be produced in amounts of 75 g/l in a fermentation process with an Arthrobacter oxydans
strain (Hahimoto et al. 1994).

Originally the strain ATCC21608 was bought from the ATCC (American Type Culture Col-
lection) because of the indication of L-methionine production (Nakayama et al. 1973). De-
spite several attempts to reproduce these experiments to obtain L-methionine, there was no
success gaining L-methionine with the material obtained from the ATCC. There are various
possible reasons: e.g. the strain material was stored at the ATCC at the beginning of the
1970s, the properties of the strain material could be altered at this long storage time.

- 66 -
___________________________________________________________________________
4.4 Corynebacterium glutamicum KY10574

4.4.1 First cultivation in the shake flask scale

The strain Corynebacterium glutamicum KY10574 was obtained from the Japanese company
Kyowa Hakko Kogyo Ltd.. The first information about cultivation techniques and suitable
media came from this company. According to this data the strain was cultivated first in a
complex medium (KH1, see Table 3.4) for 24 h. After that the minimal main medium (KH2,
see Table 3.5) was inoculated in a ratio of 1:11. This cultivation was carried out in two scales
(1 ml preculture KH1 plus 10 ml minimal medium KH2 in a 250 ml shake flask vessel; 10 ml
preculture KH1 plus 100 ml minimal medium KH2 in a 1 L shake flask vessel). The results in
both scales were comparable; in the following the results for the cultivation in the bigger scale
are presented.

0
100
200
300
400
500
600
700
L
-
M
e
t
h
i
o
n
i
n
e

L
-
A
l
a
n
i
n
e

L
-
V
a
l
i
n
e

L
-
G
l
u
t
a
m
i
c

a
c
i
d
L
-
L
y
s
i
n
e

G
l
y
c
i
n
e

L
-
L
e
u
c
i
n
e

L
-
i
s
o
-
L
e
u
c
i
n
e

A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
KH1 medium ingredients
KH1 medium 24 h cultivation

Figure 4.9: Cultivation of Corynebacterium glutamicum KY10574 in the shake flask scale in KH1
medium (100 ml medium in 1 L volume shake flask, 2 baffles, 30C, 120 rpm); directly inoculated
with a loopful of cells from BY agar plates. Presented are the amino acid concentrations before and
after the 24 h cultivation; amino acid analysis with GC.

- 67 -
___________________________________________________________________________
The precultivation in a complex medium provided increased cell growth velocity on the one
hand, but on the other side there was a major problem with the composition and the ingredi-
ents. Above all the amino acids, which were present in the complex medium in relatively high
concentration, shown in Figure 4.9 in many cases the concentration of the measured amino
acids was in many cases higher before the cultivation was started. After 24 h the concentra-
tions of 2 amino acids, L-lysine (approximately 70 mg/l) and L-valine (around 50 mg/l) were
increased in small amounts. The other amino acids had been consumed, which could be ob-
served most clearly in the case of L-glutamic acid. The reasons for consuming and new for-
mation of particular L-amino acids are very hard to establish, mostly because the composition
of complex media is only partially known.

In Figure 4.10 the results of four amino acids which were produced in KH2 medium (see Ta-
ble 3.5) with the highest concentrations are shown. The data for the other four measured
amino acids (L-leucine, L-isoleucine, L-valine, L-glutamic acid) are added in the attachment
(chapter 8). The microorganism was able to produce L-lysine in amounts around 700 mg/l in
maximum, followed by L-alanine with over 300 mg/l after 96 h. The production of glycine
and L-methionine in significant amounts started rather late (72 h); glycine was produced in
concentrations above 200 mg/l at the top; the L-methionine concentration reached concentra-
tions of 135 mg/l at the maximum after 144 h. The glucose was consumed totally after 144 h
(see Figure 4.10); all amino acid concentrations dropped down after this point in time. Excep-
tionally, L-alanine showed a fast increase and in the next step in almost the same manner a
fast decrease before the carbon source was exhausted completely.

It can be concluded from the L-alanine production data in this and chapter 4.3 that L-alanine
production was stopped as soon as the glucose level fell below certain concentration levels.
As mentioned before (see Figure 2.12 and chapter 4.3) L-alanine is degraded biochemically
through the glucogenic pathway which allows the recycling of degradation products in glu-
coneogenesis to rebuilt glucose molecules. The L-lysine concentration also decreased under
glucose limitation and as a ketogenic amino acid a conversion between L-alanine and L-lysine
is extremely unprobable. On the other hand glycine and L-methionine are glucogenic amino
acids , their concentrations increased in the L-alanine degradation phase (see Figure 4.10). L-
methionine is converted directly to succinyl-CoA in the citric acid cycle, glycine is converted
glucogenically just as alanine to pyruvate. Therefore, a conversion between these two amino
acids is the most probable case, the conversion to L-methionine is less probable. The decom-

- 68 -
___________________________________________________________________________
position of the L-alanine molecule was very fast as described and shown in Figures 4.10, 4.8
and 4.7.

The course developments for bio dry mass (BDM) and pH, shown in Figure 4.10, are similar
to those of Figure 4.7. The maximum bio dry mass value was measured after 90 h with 11 g/l.
The course for the bio dry mass was directly connected to the glucose level. The pH value
decreased during the production phase / exponentiell cell growth phase and increased during-
the stationary phase / limitation phase and in the declining phase.

L-Lysine was produced mostly in the KH2 minimal medium (see Table 3.5). It had been al-
ready produced in the preculture (see Figure 4.9), so there were distinct start conditions for
the production in the minimal medium KH2. L-Lysine was produced very fast in big amounts
after inoculation of the minimal medium KH2 (see Figure 4.10). These results indicated that
the production of L-lysine can be shifted to other amino acids (like L-methionine and glycine,
which can be detected later in the cultivation). The problem was the usage of the complex
medium for precultivation. This favoured the production of L-lysine, but the strain was not
able to grow directly on the minimal medium KH2.

- 69 -
___________________________________________________________________________

Time (h)
0 25 50 75 100 125 150 175 200 225 250
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
100
200
300
400
500
600
700
800
Methionine
Alanine
Lysine
Glycine

Time (h)
0 50 100 150 200 250
B
D
M

(
g
/
l
)
0
2
4
6
8
10
12
G
l
u
c
o
s
e

(
g
/
l
)
0
20
40
60
80
100
p
H
5,5
6,0
6,5
7,0
7,5
8,0
8,5
BDM
Glucose
pH

Figure 4.10: Amino acid production during the cultivation of Corynebacterium glutamicum KY10574
in the shake flask scale in KH2 minimal medium with 100 g/l glucose (100 ml KH2 medium in 1 L
volume shake flask, 2 baffles, 30C, 120 rpm), inoculated with 10 ml preculture KH1 (see Figure 4.9);
amino acid analysis with GC.

- 70 -
___________________________________________________________________________
The production of L-lysine with Corynebacteria strains is well known in scientific research
and industrial application (see chapter 2.3.3). It is possible to gain amounts of L-lysine with
Corynebacterium glutamicum which are more than ten times higher than the amounts ob-
tained with this strain (Eggeling et al. 1999). Moreover, in this work the target was the pro-
duction of L-methionine in amounts as high as possible. Therefore, efforts were made to pre-
vent the L-lysine production and to increase the L-methionine in the course of these altera-
tions. It was attempted to change the production pattern by the alteration of significant pa-
rameters. In detail, changes were made by altering the following conditions:

- The inoculation ratio (1:11) of the KH1 (preculture) / KH2 (minimal main medium)
was changed to 1:110
- The buffer system in the minimal medium (CaCO
3
) was replaced through morpholine
propane sulphonic acid (MOPS) in order to establish a more efficient and soluble
buffer system
- The stirring speed was increased to higher values (120 210 rpm)
- Usage of centrifuged biomass for inoculation (obtained through cultivation with KH1
complex medium)

Unfortunately all efforts were unsuccessful in terms of the enhancement of L-methionine pro-
duction. The only method to alter the production pattern in this case was the development of a
cultivation procedure without using a complex medium. The strain Corynebacterium glu-
tamicum KY10574 required just as the strain Corynebacterium glutamicum ATCC21608
more than the supply with basic nutrients to be able to grow. For this purpose, the minimal
medium F1 was developed (see Table 3.6).

4.4.1.1 Use of GC-MS amino acid identification for KH1 / KH2 media
cultivation

The cultivation as explained in the last chapter was the first successful experiment to obtain
L-methionine in higher concentrations than necessary for the survival of the bacterial strain.
Therefore, additional measurements of the samples after 96 h and 120 h cultivation time (see
Figure 4.10) were performed at the Institute of Organic Chemistry, TU Braunschweig. It was
possible to get an additional and secure proof of the accuracy of the results shown in chapter
4.4.1. The identification of the modified amino acids could be done with GC separation and

- 71 -
___________________________________________________________________________
following mass spectrometry (MS) analysis which allowed a precise identification of the sub-
stances.

The first step for the identification is the exact knowledge of the derivatization process of the
amino acids, performed with a GC kit for free physiological amino acids (see chapter 3.4.7).
The free information from the company Phenomenex about this chemical process was not
completed, as shown in Figure 4.11.

O
OH N H
2
S C H
3
+ 2
Cl O
O
O
O NH
O
O
S C H
3
+ 2 HCl + CO
2
L-methionine
R
R
R

Figure 4.11: Chemical modification of amino acids for instance of L-methionine by derivatization due
to the information of the company Phenomenex (used GC kit: EZ faast for free physiological amino
acids).

In order to get additional information about the nonspecified side chains (R) and the use of
catalysts, the derivatization agents 1 to 6 (see Table 3.10) were investigated by GC-MS. It
could be shown that the important chemicals for the derivatization process were contents of
the reagents 3A and 4. In reagent 3B the catalysator 3-methyl pyridine was identified; in re-
agent 4 the substance propylchloroformate was identified which indicated that the unnamed
side chains in Figure 4.11 stand for propyl groups. The original measurement data of the re-
agents 4 and 3B are added to the attachment (chapter 8). The completed chemical reaction for
the derivatization of amino acids is shown in Figure 4.12. Several scientific reports have been
published about amino acid analysis and derivatization before measurement. The chemical
reaction mechanism shown in Figure 4.12 is also described in principle from Husek (Husek et
al. 1990). With the information about the amino acid modification it was possible to identify
the fragments received from the mass spectrometry.

- 72 -
___________________________________________________________________________
O
OH N H
2
S C H
3
+ 2
Cl O
O
CH
2
C H
2
CH
3
N
CH
3
Catalysator
O
O NH
O
O
C
H
2
C
H
2
C H
3
C
H
2
C
H
2
CH
3
S C H
3
+ 2 HCl + CO
2
L-methionine
propylchloro -
formate
3-methyl
pyridine

277 g/mol

Figure 4.12: Completed chemical reaction for the modification of amino acids by derivatization be-
fore GC measurement for the instance of L-methionine after determination with GC-MS

In the next step the sample of the cultivation of Corynebacterium glutamicum KY10574 after
120 h (see Figure 4.10) was measured with GC-MS at the Institute of Organic Chemistry, TU
Braunschweig. The sample was modified through derivatization as shown in Figure 4.12 and
in the same way as all samples obtained from the cultivations. The results obtained from the
GC measurement at the Institute of Organic Chemistry are shown in Figure 4.13. In addition
Figure 4.14 shows the chromatographic patterns obtained with the gas chromatography at the
Institute of Biochemistry and Biotechnology in order to be able to compare the results (see
next page for Figures 4.13 + 4.14).

The chromatographic patterns in Figures 4.13 and 4.14 are similar, but the retention times are
different. This can be referred to the different length of the used GC columns, but the material
of the columns (solid stationary phase) were identical. For this reason the interactions be-
tween the samples and the GC column material were similar. The chromatogram in Figure
4.14 shows more background peaks at the base line than in the comparable chromatogram in
Figure 4.13 at which a very clear baseline can be recognised.

- 73 -
___________________________________________________________________________

Valine
Glutamic acid
iso-Leucine
Methionine
Lysine
Leucine
Norvaline
200 mol/l
Glycine
Alanine

Figure 4.13: Gas chromatography (GC) results after derivatization of the cultivation sample after 120
h in the shake flask scale (see Figure 4.10); this experiment was performed at the Institute of Organic
Chemistry, TU Braunschweig. Gas chromatograph type Agilent 6890, equipped with a 30 m analytical
column (Phenomenex ZB5-MS, 30 m 0,25 mm ID).

Alanine
Glycine
Valine
Norvaline;200 mol/L
internal standard
Leucine
Iso-leucine
Methionine
Glutamic acid
Lysine

Figure 4.14: GC results after derivatization of the cultivation sample after 120 h in the shake flask
scale (see Figure 4.10); this experiment was performed at the Institute of Biochemistry and Biotech-
nology, TU Braunschweig. Gas chromatograph type Hewlett Packard HP5890, equipped with column:
Chrompack CP-Sil-8CB, low bleed capillary (15 m 0.25 mm 0.25 m).

- 74 -
___________________________________________________________________________
In Figures 4.13 + 4.14 some chromatographic peaks cannot be allocated precisely to a distinct
amino acid or, in general, a definite chemical compound. These unidentified compounds have
to hold an amino group in their chemical structure. This is referred to the derivatization proc-
ess before measurement which filtered out all substances without those properties. On the
other side this method allowed to identify 8 amino acids and the internal standard L-norvaline
without any doubt, among them the most important metabolites (above all L-methionine, L-
lysine, L-glutamic acid, glycine and L-alanine) which can be produced with Corynebacterium
glutamicum bacteria. The fact, that the strain Corynebacterium glutamicum KY10574 was
able to produce such a broad amino acid spectrum in different media proved the amino acid
production variety of this strain under the described conditions.

The metabolism of L-methionine, L-lysine, L-alanine and glycine was deregulated, otherwise
it could not be explained why the microorganism was able to produce these amino acids in
higher values than necessary for its own supply with nutrients. Under these considerations the
composition of the medium becomes more and more important, because the microorganism
seems to be able to choose the metabolic production pathway depending on the offered nutri-
ents from the available medium.

The deregulation of several amino acid biochemical pathways occurred by randomised
mutagenesis. This microbiological method can improve the abilities of microorganisms to
overproduce metabolites considerably (see chapter 2.5) but generates also, apart from insert-
ing the favoured mutation (e.g. the derugulation of L-methionine feedback repression and
inhibition), simultaneous inserts which lead to unrequested additional mutations with un-
wanted side effects.

The Figure 4.15 shows the identification of the amino acid L-methionine with mass spec-
trometry (MS). The amino acids were separated due to the GC method (see Figure 4.13 +
4.14), so it was possible to analyze them in a MS system. The amino acids, in this case L-
methionine, were degraded by electronic ionisation. The obtained fragments from a methion-
ine reference and a real sample are presented in Figure 4.15. The Figure presents a time of
flight (TOF) graph, the fragments were detected in distinct points in time.

- 75 -
___________________________________________________________________________

propyl
fragment
molecular ion peak
reference
methionine
(modified)
sample from culti-
vation 120 h
Figure 4.15: Mass spectrometry (MS) results for the example of modified L-methionine (see Figure
4.15) from a reference sample (at the bottom) and a cultivation sample after 120 h (at the top); see for
the cultivation Figure 4.10. This experiment was performed at the Institute of Organic Chemistry, TU
Braunschweig with a time of flight mass spectrometer type JMS-T100GC (GCAccuTOF, JEOL, Ja-
pan).

Each chemical compound shows significant fragment patterns after mass spectrometry (MS)
analysis including molecular ions and ions from certain degradation steps. The modified form
of L-methionine has a molar mass of 277 g/mol (see Figure 4.15). In the MS-TOF graph of
both samples the residual unfragmented molecule can be detected (277 g/mol). Another sig-
nificant fragment is the first one in both diagrams in Figure 4.15, the propyl fragment. As
shown in Figure 4.12, 2 propyl groups are connected to L-methionine in the derivatization
process; so the existence of these groups is another important information for the interpreta-
tion of the diagram. Nowadays, large databases exist to identify chemical substance pattern
gained from MS. The mass spectrometry diagrams for the other identified amino acids (L-
alanine, L-valine, L-glutamic acid, L-lysine, L-leucine, L-isoleucine, glycine) are added to the
attachment (chapter 8).

- 76 -
___________________________________________________________________________
4.4.2 Shake flask cultivation of Corynebacterium glutamicum
KY10574 in F1 minimal medium

After first attempts to cultivate the strain Corynebacterium glutamicum KY10574 in a combi-
nation of complex medium (KH1, see Table 3.4) as preculture and a minimal medium (KH2,
see Table 3.5) for production of the amino acid L-methionine, it became evident, that it was
impossible to enhance the L-methionine production this way. Furthermore, the production of
methionine began very late in the main culture (KH2 medium), after approximately 80 90 h,
not including the 24 h of precultivation. Another problem was the carbonate buffer system
(CaCO
3
), which was not very efficient. It complicated the measurement of the BDM and the
optical density (OD).

The strain Corynebacterium glutamicum KY10574 had problems growing on pure minimal
media; it was unable to grow on minimal medium SMM (Table 3.2) and KH2 (Table 3.5).
Therefore, the developed minimal medium F1 (see Table 3.6) was tested. It contained a broad
variety of mineral elements and vitamins. The first cultivation of C. glutamicum KY10574 in
F1 minimal medium is shown in Figure 4.16.
Time (h)
0 20 40 60 80 100 120 140 160 180
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
50
100
150
200
250
300
Methionine
Alanine
Lysine
Glycine

Figure 4.16: Amino acid production during cultivation of C. glutamicum KY10574 in F1 minimal
medium, 20 g/l glucose, 30C, 150 rpm, 25 ml medium in a 250 ml shake flask vessel (3 baffles);
inoculated directly from BY agar plates; amino acid analysis with GC.

- 77 -
___________________________________________________________________________
The change of the amino acid product spectrum compared to former cultivations (Figure 4.10)
is obvious. There was only a small amount of L-lysine, which was the dominant amino acid in
the cultivations with KH1 + KH2 media. The amounts of glycine are comparable to those
shown in Figure 4.10. However, it lasted much longer in the former cultivations with complex
precultures until the production achieved the same concentrations as in this cultivation. The
production of L-methionine began to start very rapidly and stopped after 66.5 h at with 286
mg/l. The concentration and the production of L-methionine were more than doubled com-
pared to Figure 4.10. The slight decrease of concentration at the end of the shake flask culti-
vation can be explained by glucose limitation. In general, the microorganisms are stopping the
metabolite production or they alter the production to other and easier constructed molecules.
In this case (Figure 4.16) a slight product shift from L-methionine to L-lysine after the com-
plete consumption of the carbon source could be observed. L-alanine was produced very late
and only in small amounts. In the following chapter 4.4.3 it is pointed out that the L-alanine
production started to increase with the presence of higher glucose levels in the medium.

Another important experience from this cultivation was the influence of the shaking rate. In
the first attempts with a shaking rate of 120 rpm, the strain began either to grow only after a
long adaptation phase or there was actually no cell growth at all. This is remarkable because
the cultivations were performed at the beginning only in small scales with 25 ml minimal me-
dium F1 in a 250 ml shake flask scale. The inoculation was done with a loopfull of cells from
BY agar plates. With higher shaking rates (150 rpm) the time of the lag-phase was shortened
for the benefit of the exponentiell growth phase. The strain C. glutamicum KY10574 and in
general all C. glutamicum strains exhibited a high demand for oxygen (Gomes et al. 2005;
Sharma et al. 2001). According to this data it was important to provide good aeration condi-
tions by intensive mixing through the shaking rate and through the baffles in the shake flask
walls to rip open the surface of the cultivation broth. By this procedure there was an enlarge-
ment of the liquid surface which allowed a much higher oygen uptake rate (OUR) inside the
liquid medium. Especially at the beginning good aeration was absolutely crucial, since the
modality of cell growth depended at first on optimised mixing conditions and oxygen supply.
A real optimisation of the aeration could be better performed in bioreactors with controlled
gassing in combination with an effective stirrer speed for mixing which was much more effi-
cient than the shaking rates in the shake flask scale. This is explained in detail in the bioreac-
tor cultivations with the strain Corynebacterium glutamicum KY10574 in chapter 4.4.7.

- 78 -
___________________________________________________________________________
A sample, taken from the foregoing cultivation from Figure 4.17 after 66.5 h (maximum L-
methionine production) was used as inoculum for the shake flask cultivation shown in Figure
4.17. The target was the enhancement of L-methionine production by transferring a preculture
with high methionine concentration in a fresh medium. Due to this preculture the amino acid
concentrations increased practically without a noticeable adaptation (Figure 4.17). The L-
methionine maximum was reached very early after 27 h. After about 50 h most of the amino
acid concentrations began to decrease. The L-methionine concentration was only slightly en-
hanced by 50 mg/l from 286 mg/l after 66.5 h (Figure 4.16) to 333 mg/l and 339 mg/l after 27
h respectively 51 h (Figure 4.17). The concentration of the other amino acids remained un-
changed. So it was not very effective to increase the L-methionine yield using samples as in-
oculum like it was done in this case. The level of glucose and the glucose limitation at a dis-
tinct point in the cultivation process seemed to be the more crucial part in order to rise the
amino acid formation and production.
Time (h)
0 20 40 60 80 100 120
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
50
100
150
200
250
300
350
400
Methionine
Alanine
Lysine
Glycine

medium, 20 g/l glucose, 30C, 150 rpm, 25 ml medium in 250 ml shake flask vessel (3 baffles); inocu-
lated with preculture: 2.5 ml sample taken from the cultivation shown in Figure 4.16 after 66.5 h;

- 79 -
___________________________________________________________________________
To be sure that the GC-detection of methionine is in qualitative agreement with well-known
colour-based methods, routinely thin layer chromatography (TLC) was used. The thin layer
chromatography of 2 samples obtained from the shake flask cultivation, shown in Figure 4.17
(27 h and 51 h), reveals the results from the gas chromatography in a visual way. The reten-
tion times of the methionine spots in both samples was completely identical with the that of
the methionine reference which served as an additional proof for the correctness of the meas-
urements. In comparison to the reference it was also evident that an analysis of amino acid
concentration by means of this method can be helpful to get index values, but not for quantita-
tive determinations of the amino acid concentrations.

Methionine
400 mg/l
Sample
339 mg/l
Sample
333 mg/l

Figure 4.18: Qualitative thin layer chromatography (TLC) of the samples after 27 h and 51 h from the
cultivation shown in Figure 4.17 and a reference of L-methionine (400 mg/l); detection with ninhy-
drin.

- 80 -
___________________________________________________________________________
4.4.3 Shake flask cultivations in F1 minimal medium with different
glucose levels at rising shaking rates

After first positive results obtained with the minimal medium F1 in the shake flask scale sys-
tematic investigations were done to improve the L-methionine production conditions. The
first experimental series were performed to optimise the glucose level of the F1 medium and
the shaking rate. The glucose level varied between 20 and 100 g/l, the shaking rates between
150 and 210 rpm. It was the target to adjust the highest possible shaking rate to enhance the
oxygen transport and supply in the shake flask cultures, which seems to be very important for
both a cell growth and metabolite production of Corynebacterium glutamicum (Sharma et al.
2001). Due to the increasing density of the cultivation liquid in the course of the cultivation
time (caused by the bio dry mass development) the shaking rates were boosted gradually in
the following steps: 150 rpm (0 45 h), 180 rpm (ca. 45 72 h), 210 rpm ( 72 h). These
alterations were carried out for both scales: 25 ml medium in 250 ml shake flask and 100 ml
medium in 1 L shake flask. In the following chapters there is a complete exemplary presenta-
tion of the results for the small scale (Figures 4.19 4.23), an example for the big scale culti-
vation (Figure 4.24) and a comparison of the measurements from both scales (Figure 4.25).
The complete series is shown to give clear information about the direct context between the
amino acid production on the one hand and the cell growth development as well as the glu-
cose consumption on the other hand.

- 81 -
___________________________________________________________________________
4.4.3.1 25 ml cultivation (250 ml shake flask) with 20 g/l glucose

The experimental series was started with 20 g/l of glucose. The L-methionine and glycine
production values shown in Figure 4.19 can be directly compared with the first cultivations
with this medium (F1) as presented in Figure 4.16. The variations in the product values for L-
methionine and also for glycine were in comparison of both diagrams in normal ranges. L-
methionine concentrations of approximately 230 mg/l after 44 h can be achieved with a glu-
cose level of 20 g/l, glycine amounts were around 200 mg/l. Other amino acids were produced
only in small amounts (< 50 mg/l) as shown for L-alanine and L-valine in Figure 4.19.

The decrease of the L-methionine concentration after 50h occurred probably due to the fast
consumption of glucose (consumed after 30 h). As described before amino acids like alanine,
methionine and glycine are degraded to products of the citric acid cycle leading at the end to
gluconeogenesis (see Figure 2.12 and chapter 4.3). Therefore, it can be assumed that these
amino acids are degraded in order to supply new glucose molecules for maintaining the mi-
crobial metabolism. These results suggest further investigations with higher levels of glucose.

The biomass reached 7 g/l after 30 h, further cell growth was not possible due to the con-
sumed carbon source (glucose). As expected the development of the course for the optical
density was parallel with the bio dry mass.

- 82 -
___________________________________________________________________________
Time (h)
0 20 40 60 80 100 120
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
25
50
75
100
125
150
175
200
225
250
275
300
Methionine
Glycine
Alanine
Valine

Time (h)
0 20 40 60 80 100 120
G
l
u
c
o
s
e

(
g
/
l
)
0
5
10
15
20
25
B
D
M

(
g
/
l
)
0
1
2
3
4
5
6
7
8
O
D
5
4
6
0
5
10
15
20
25
30
35
40
Glucose
BDM
OD
546

Figure 4.19: Amino acid production and development of bio dry mass (BDM), optical density and
glucose (g/l) during cultivation of C. glutamicum KY10574 in F1 minimal medium, 20 g/l glucose,
30C, 25 ml medium in a 250 ml shake flask vessel (3 baffles); inoculated directly from BY agar
plates; shaking rate: 150 rpm (0 45 h), 180 rpm (ca. 45 72 h), 210 rpm ( 72 h); amino acid analy-
sis with GC.

- 83 -
___________________________________________________________________________

The cultivation results with 40 g/l of glucose in the F1 minimal medium are shown in Figure
4.20. The L-methionine concentration reached levels above 325 mg/l after 45 h before the
amounts dropped quickly. In certain cases higher L-methionine production could be observed,
but in most experiments with this glucose concentration in the medium there were concentra-
tions around 300 400 mg/l L-methionine in the cultivation broth. But the exceptions with
higher L-methionine amounts in the shake flasks proved a capability for advanced production.

The glucose was again consumed very rapidly due to attempts to provide high oxygen supply
into the medium. The bio dry mass values were doubled compared to the experiment shown in
Figure 4.19, but this fact did not yet successfully increase the L-methionine production sig-
nificantly. The production of glycine is increased slightly from 200 mg/l (Figure 4.19) to val-
ues of 250 mg/l in Figure 4.20. The L-valine concentrations are on low levels, only after 30 h
a value above 50 mg/l could be measured.

The L-alanine concentration rose quickly after cultivation had started but dropped in the same
way immediately after the consumption of the carbon source. The instantaneous decrease of
L-alanine concentrations under glucose limited conditions was characteristical for this amino
acid, no other amino acid reacted in this way with the glucose level. This is most probably
due to biochemical degradation in order to gain new glucose molecules through gluconeo-
genesis (Figure 2.12, chapter 4.3, chapter 4.4.3.1).

- 84 -
___________________________________________________________________________
Time (h)
0 20 40 60 80 100 120
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
50
100
150
200
250
300
350
Methionine
Glycine
Alanine
Valine

Time (h)
0 10 20 30 40 50 60 70 80 90 100 110 120
G
l
u
c
o
s
e

(
g
/
l
)
0
5
10
15
20
25
30
35
40
45
50
B
D
M

(
g
/
l
)
0
2
4
6
8
10
12
14
16
18
O
D
5
4
6
0
10
20
30
40
50
60
70
80
Glucose
BDM
OD
546

glucose (g/l) during cultivation of C. glutamicum KY10574 in F1 minimal medium, 40 g/l glucose,
30C, 25 ml medium in a 250 ml shake flask vessel (3 baffles); inoculated directly from BY agar
plates; shaking rate: 150 rpm (0 45 h), 180 rpm (ca. 45 72 h), 210 rpm ( 72 h); amino acid analy-
sis with GC.

- 85 -
___________________________________________________________________________

The cultivation results with 60 g/l of glucose in the F1 minimal medium are shown in Figure
4.21. Both the bio dry mass development and the amino acid production were successfully
increased, especially the L-methionine concentration increased considerably. After 50 h more
than 700 mg/l were produced and even in the following time under glucose limited conditions
(see Figure 4.21) the concentration increased to values near 900 mg/l after 120 h. The reason
could be the release of intracellular accumulated L-methionine in the supernatant after the cell
wall collapse due to nutrient deficiency. This effect could also be observed for L-valine
(maximum value: 300 mg/l after 120 h) and glycine (maximum value: 460 mg/l after 120 h).
The course development for L-alanine was different, the maximum production was located
after 46 h with 300 mg/l; afterwards the characteristical decrease could be observed. The
maximum values of all amino acids shown in Figure 4.21 increased compared to the cultiva-
tion experiments with lower glucose levels at start (Figure 4.20 and 4.19).

The bio dry mass reached levels higher than 14 g/l. In contrast to the cultivation with 40 g/l
glucose in the medium a very efficient transfer of enhanced cell growth (which was reached
through the increase from 20 to 40 g/l of glucose in the medium) to a noticeable enhancement
of the observed production of all amino acids could be seen.

- 86 -
___________________________________________________________________________
Time (h)
0 20 40 60 80 100 120
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
100
200
300
400
500
600
700
800
900
1000
Methionine
Glycine
Alanine
Valine

Time (h)
0 20 40 60 80 100 120
G
l
u
c
o
s
e

(
g
/
l
)
0
10
20
30
40
50
60
B
D
M

(
g
/
l
)
0
2
4
6
8
10
12
14
16
18
O
D

5
4
6
0
10
20
30
40
50
60
70
Glucose
BDM
OD
546

glucose (g/l) during cultivation of C. glutamicum KY10574 in F1 minimal medium, 53 g/l glucose
(value by analysis), 30C, 25 ml medium in a 250 ml shake flask vessel (3 baffles); inoculated directly
from BY agar plates; shaking rate: 150 rpm (0 45 h), 180 rpm (ca. 45 72 h), 210 rpm ( 72 h);

- 87 -
___________________________________________________________________________

The cultivation with a glucose content of 80 g/l at the beginning featured surprising results, as
shown in Figure 4.21. As it has been indicated before the L-alanine concentration increased
with higher glucose levels but in this case the production of L-alanine came to pass to the dis-
advantage of the amino acids L-methionine and glycine (compare with Figure 4.21). L-
alanine was obtained in amounts of approximately 600 mg/l without further concentration
decrease (Figure 4.30). L-methionine was gained in amounts of nearly 500 mg/l, glycine con-
centrations achieved levels of ca. 250 mg/l. Another difference was the appearance of L-
glutamic acid at 300 400 mg/l. This amino acid was produced at very low amounts in for-
mer cultivations using lower glucose concentrations in the medium. With a constant concen-
tration around 300 mg/l L-valine was the sole exception. Compared to Figure 4.21 no obvious
fluctuation took place.

The glucose was not consumed before 72 h due to the high provided concentration at the be-
ginning and the values for the BDM (approx. 15 - 16 g/l) and the OD
546
were in normal
ranges.


The results obtained with 100 g/l glucose in F1 minimal medium are shown in Figure 4.23.
The amino acid production patterns which were obtained in this case were similar to those in
the experiments with 80 g/l. There were only minor alterations in the product amounts com-
pared to Figure 4.22. The highest production of any amino acids is L-alanine with 625 mg/l at
the maximum, followed by L-methionine (ca. 500 mg/l), L-valine (ca. 330 mg/l), glycine (ca.
270 mg/l) and L-glutamic acid (ca. 235 mg/l). The L-methionine production finished after 47
h. The glucose was not consumed totally and the values settled down after 100 h at 10 g/l.
This indicated a glucose excess supply and the ocurrence of the limitation of other nutrient
components (e.g. nitrogen, sulphur or phosphate) in the cultivation. L-alanine was not de-
graded to intermediates of the citric acid cycle; due to the glucose excess there was no neces-
sity to form new glucose molecules from amino acids for the metabolism. In accordance with
that the values for biomass and OD
546
finished on a level known from the cultivation reported
before.

- 88 -
___________________________________________________________________________
Time (h)
0 20 40 60 80 100 120
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
100
200
300
400
500
600
700
800
Methionine
Glycine
Alanine
Valine
Glutamic acid

Time (h)
0 20 40 60 80 100 120
G
l
u
c
o
s
e

(
g
/
l
)
0
10
20
30
40
50
60
70
80
B
D
M

(
g
/
l
)
0
2
4
6
8
10
12
14
16
18
O
D
5
4
6
0
10
20
30
40
50
60
70
80
Glucose
BDM
OD
546

glucose (g/l) during cultivation of C. glutamicum KY10574 in F1 minimal medium, 73.5 g/l glucose

- 89 -
___________________________________________________________________________
Time (h)
0 20 40 60 80 100 120
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
100
200
300
400
500
600
700
Methionine
Glycine
Alanine
Valine
Glutamic acid

Time (h)
0 20 40 60 80 100 120
G
l
u
c
o
s
e

(
g
/
l
)
0
10
20
30
40
50
60
70
80
90
100
B
D
M

(
g
/
l
)
0
5
10
15
20
O
D
5
4
6
0
10
20
30
40
50
60
70
80
Glucose
BDM
OD
546

glucose (g/l) during cultivation of C. glutamicum KY10574 in F1 minimal medium, 95 g/l glucose

- 90 -
___________________________________________________________________________
4.4.3.6 100 ml cultivation (1 L shake flask) with 60 g/l glucose

The highest amount of L-methionine in the 250 ml shake flask scale (3 baffles) with 25 ml F1
minimal medium was gained with 60 g/l glucose as carbon source at the beginning of the cul-
tivations. All experiments were also performed with 1 L shake flasks (2 baffles) and 100 ml
medium. Also, the shaking rates were altered in the same way: 150 rpm (0 45 h), 180 rpm
(ca. 45 72 h), 210 rpm (since ca. 72 h). The comparable results for the cultivations with a
glucose level of 53 g/l with 100 ml medium in a 1 L shake flask are shown in Figure 4.24.
The differences become clear by comparing the produced amounts of L-methionine as shown
in Figure 4.24 with the amounts shown in Figure 4.21 using the smaller cultivation scale. L-
Methionine was produced in maximum of 400 mg/l after 96 h, the other amino acids were
also diminished in their concentration values.
Time (h)
0 20 40 60 80 100 120 140 160
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
100
200
300
400
500
Methionine
Alanine
Valine
Glycine

medium, 53 g/l glucose (value by analysis), 30C, 100 ml medium in a 1 L shake flask vessel (2 baf-
fles); inoculated directly from BY agar plates; shaking rate: 150 rpm (0 45 h), 180 rpm (ca. 45 72
h), 210 rpm ( 72 h); amino acid analysis with GC.

It was obvious that there were significantly lower production levels of amino acids in all ex-
periments with the bigger scale cultivation. The reason for that was most probably the insuffi-
cient oxygen supply which could not be guaranteed in this shake flask scale using 100 ml me-

- 91 -
___________________________________________________________________________
dium. Another reason for reduced mixing and oxygen supply could be also the lower number
of baffles (2) in 1 the 1 L shake flasks compared to the 250 ml shake flasks (3). Consequently,
C. glutamicum did not produce the amino acids derived in the lower scale experiments (Go-
mes et al. 2005). The major influence of oxygen supply on amino acid production with
C.glutamicum KY10574 was also clearly visible in the bioreactor experiments (chapter 4.4.6).
Therefore, there is a comparison of the results obtained from both scales in Figure 4.25.

4.4.3.7 Comparison of the L-methionine production with C.glutamicum
KY10574 with different glucose amounts and shake flask scales

The completion of chapter 4.4.3 with its series of experiments on the influence of the carbon
source (glucose) and the rising shaking rates (150 210 rpm) in 2 different shake flask scales
is presented in Figure 4.25. The highest level of L-methionine production (900 mg/l after 119
h) could be observed with 60 g/l glucose at the beginning of the smaller cultivation scale.
Concentrations around 500 mg/l of L-methionine were obtained in the small shake flask scale
with glucose concentrations around 60 g/l. It was not expedient to use smaller amounts than
40 g/l in the medium because of the fast glucose consumption and subsequently the end of
cell growth due to carbon source limitation. In the bigger shake flask scale the L-methionine
level reached values which were only near 500 mg/l at the maximum (using 60 + 80 g/l glu-
cose in medium). Obviously, there was a limitation in oxygen supply but there was no way to
increase the shaking rate speed above 210 rpm due to the massive splashing of the liquid at
these velocities. A problem in all shake flasks was a substantial layer of foam on the surface
of the cultivation broth which impeded the transport of oxygen in the interior areas of the liq-
uid. The foam was also one of the biggest problems in the bioreactor cultivations (chapter
4.4.7).

The results from Figure 4.25 represent a summary of an exemplary experimental series (Fig-
ure 4.19 4.23) which is very characteristical for amino acid production with Corynebacte-
rium glutamicum KY10574. But the experiments were carried out with the same conditions
several times. The average of L-methionine production values in all shake flasks which were
performed under the same conditions with shaking rates from 150 210 rpm and different
glucose amounts are shown in Figure 4.26. There are some slight alterations in the L-
methionine production level compared to Figure 4.25. The highest concentrations were
reached using 60 g/l of glucose as carbon source in the small scale (25 ml medium in 250 ml

- 92 -
___________________________________________________________________________
shake flasks) and 80 g/l of glucose in the big scale (100 ml medium in 1 L shake flasks). In
general, the L-methionine values were higher using glucose values between 70 and 100 g/l
than values between 20 and 40 g/l. Glucose levels under 50 g/l were consumed very quickly
and caused an early amino acid production stop. The product spectrum of amino acids can be
altered through minor changes in parameters like cell growth, pH and shaking rates and me-
dium composition. Additionally the strain C. glutamicum KY10574 is able to produce a broad
variety of amino acids due to its setting up through random mutagenesis by Kyowa Hakko.
Therefore, marginally alterations, e.g. in the important phase at the beginning of the cell
growth might change considerably the development of the product spectrum in each shake
flask.

0
100
200
300
400
500
600
700
800
900
1000
100 80 60 40 20
M
e
t
h
i
o
n
i
n
e

(
m
g
/
l
)
25 ml culture (250 ml shake flask)
100 ml culture (1 L shake flask)

Glucose (g/l)

47.5 h 47.5 h 119 h (0.25 L)
96 h (1L)
96 h (0.25 L)
118.5 h (1L)
118.5 h

Figure 4.25: Comparison of the cultivations seen before on the influence of glucose amounts and 2
different shake flask scales (25 ml F1 medium in 250 ml shake flasks (3 baffles); 100 ml F1 medium
in 1 L shake flasks (2 baffles) on the L-methionine production capability of C. glutamicum KY10574;
shaking rate mode: 150 210 rpm (180 rpm since 45 h; 210 rpm since 72 h); temperature 30C; amino
acid analysis with GC. (The points in time when the L-methionine values were measured are indi-
cated).

- 93 -
___________________________________________________________________________
0
100
200
300
400
500
600
700
800
900
1000
20 40 60 80 100
Glucose (g/l)
M
e
t
h
i
o
n
i
n
e

(
m
g
/
l
)

Figure 4.26: Average L-methionine production values in 2 shake flask scales in F1 minimal medium
with varied glucose concentrations; 25 ml F1 medium in 250 ml shake flasks (3 baffles); 100 ml F1
medium in 1 L shake flasks (2baffles); temperature 30C; shaking rate mode: 150 210 rpm (180 rpm
since 45 h; 210 rpm since 72 h), amino acid analysis with GC.

After comparing the shake flask results in both scales the wide variety of produced amino
acids in several concentration ranges became more and more obvious. The strain C. glu-
tamicum KY10574 was derived through random mutagenesis experiments (see chapter 2.5).
A characteristic feature of this method is the appearance of alterations in several biochemical
pathways, not only in the favourised target pathways. This attribute leads inevitably to the
deregulation of several production pathways, subsequently the feedback repression and inhi-
bition signals of many amino acids are also partially deactivated. So the strain KY10574 was
developed in order to get a methionine overproducer, but other mutations could not be
avoided which lead to the production of other amino acids as by-products after the alteration
of their regulation mechanisms (Kumar et al. 2005). The directed mutation of the biochemical
pathways of amino acids in C. glutamicum by genetic engineering has the major advantage
that unrequested mutations in other biochemical pathways than the desired one can be
avoided completely (Wittmann et al. 2004).

- 94 -
___________________________________________________________________________
4.4.4 Shake flask cultivation of C. glutamicum KY10574 in F1 me-
dium with different shaking rates

The Figure 4.27 shows the differences in L-methionine production using several shaking
rates. The highest possible shaking rate at the beginning could be 120 or 150 rpm due to the
extensive splashing of the liquid medium. The shaking rate could be enhanced after 45 h (180
rpm) and 72 h (210 rpm), because the density and viscosity of the medium was much higher
due to the accumulation of biomass. The enhancement of the shaking rate in order to enhance
the L-methionine concentration was successful for the smaller scale (25 ml in 250 ml flasks).
Velocities of 150 rpm were absolutely necessary to get higher L-methionine production con-
centrations than 500 mg/l. The improvement of the oxygen supply using a shaking rate of 150
rpm was clearly visible for the smaller scale. Higher shaking rates could increase the produc-
tion only marginally. In the case of the bigger scale higher shaking rates showed no influence.
Obviously it was much more difficult to supply the microorganisms with sufficient amounts
of oxygen in this shake flask scale.

0
100
200
300
400
500
600
700
800
900
120 150 150 - 180 150 - 210
shaking rate (1/min)
M
e
t
h
i
o
n
i
n
e

(
m
g
/
l
)

Figure 4.27: L-methionine production values in 2 shake flask scales using different shaking rates
from 120 210 rpm (25 ml F1 medium in 250 ml shake flasks (3 baffles); 100 ml F1 medium in 1 L
shake flasks (2 baffles)); temperature 30C; glucose 60 g/l, amino acid analysis with GC.

- 95 -
___________________________________________________________________________
4.4.5 Shake flask cultivation in F1 minimal medium using precul-
tures

The adaptation of the microrganisms to the minimal medium F1 was an important factor. This
was a major problem in the first cultivations; in every second shake flask cultivation the mi-
croorganisms had problems to grow. In order to avoid the long transition phase for biomass
growth at the beginning and to enhance the amino acid production precultures were used.

Table 4.1 shows the concentrations of 4 amino acids after 48 h of shake flask cultivation. The
glucose concentration (56 g/l) was optimised according to the results obtained in former ex-
periments (see Figure 4.25), as well as the shaking rate mode (see Figure 4.27). This cultiva-
tion broth was the preculture for 2 shake flask experiments with the same minimal medium
(Figure 4.28, inoculated in a ratio of 1:10). These shake flask experiments were carried out
with the same conditions as in the preculture. In both shake flask experiments, which are pre-
sented in Figure 4.28, the L-methionine formation increased remarkably to values above 1.2
g/l in the first 42 h of the cultivations. There was no lag-phase for microbial adaptation at the
beginning of the cultivation after inoculation. So the microorganisms could use the available
nutrients in a very fast and efficient way which allowed a production enhancement to 1.4 g/l
(average value of both cultivations) after 144 h.

It was also noticeable that the production of L-methionine even at the end of the cultivation
time still increase a little bit which could be also observed in Figure 4.21. A possible reason
might be that at this point in time the microbial cell walls began to collapse caused by nutrient
deficiency and released intracellular amounts of L-methionine (accumulated in the cytosol)
into the extracellular area. This could promote the L-methionine production even under insuf-
ficient nutrient supply conditions.

Table 4.1: Amino acid production during cultivation of C. glutamicum KY10574 in F1 medium, 56
g/l glucose, 30C, 25 ml medium in a 250 ml shake flask vessel (3 baffles); inoculated directly from
BY agar plates; shaking rate mode: 150 rpm (0 45 h), 180 rpm; amino acid analysis with GC. This
culture was used as preculture for the cultivation shown in Figure 4.28.

Time (h) L-Methionine (mg/l) L-Alanine (mg/l) Glycine (mg/l) L-Valine (mg/l)
27 25 16.3 16.4 4.6
48 658.2 355.3 280.3 563

- 96 -
___________________________________________________________________________
Time (h)
0 20 40 60 80 100 120 140 160
L
-
M
e
t
h
i
o
n
i
n
e

(
m
g
/
l
)
0
200
400
600
800
1000
1200
1400
1600
Methionine (Cult.A)
Methionine (Cult.B)

Figure 4.28: Amino acid production during cultivation of C. glutamicum KY10574 in F1 medium, 56
g/l glucose, 30C, 25 ml medium in a 250 ml shake flask vessel (3 baffles); inoculated with 2.5 ml
from preculture (shown in Table 4.1); shaking rate mode: 150 rpm (0 45 h), 180 rpm; amino acid
analysis with GC.

- 97 -
___________________________________________________________________________
4.4.6 Shake flask cultivation of Corynebacterium glutamicum
KY10574 using metabolic precursors of L-methionine

The influence of the metabolic precursors of L-methionine on the production of the amino
acid with the strain C. glutamicum DSM20300, L-homoserine and L-homocysteine, were
shown before in chapter 4.2.2 (Figure 4.4). Homocysteine is the last metabolic precursor on
the biochemical pathway to L-methionine formation (see Figure 2.11), whereas homoserine is
a corporate precursor of the amino acids lysine, threonine and methionine (Voet et al. 2002).
The influence of the addition of 0.5 g/l of these precursors into the cultivation medium F1 was
investigated as shown in Figure 4.29 comparing it to a cultivation without these precursors.
According to these data it can be concluded that no improvement of the L-methionine produc-
tion could be observed. On the contrary, the cultivation without any precursors showed the
maximum L-methionine production.
Time (h)
0 20 40 60 80 100 120 140 160
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
100
200
300
400
500
without precursor
0.5 g/l Homoserine
0.5 g/l Homocysteine

Figure 4.29: Comparison of C. glutamicum KY10574 cultivations with and without the addition of
0.5 g/l homoserine and homocysteine in F1 minimal medium with 40 g/l glucose; 100 ml medium in 1
L shake flasks (2 baffles), directly inoculated from BY agar plates; 30C; shaking rate: 150 rpm (0
45 h), 180 rpm; 53 g/l glucose; amino acid analysis with GC.

- 98 -
___________________________________________________________________________
The inhibition mechanisms maintaining feedback repression and inhibition phenomenons
(Figure 2.10) which are responsible for deactivating the L-methionine pathway through the
addition of high amounts of the precursors could not be prevented successfully. The effect
seemed to bring about quite the contrary. The control mechanisms, which were altered in the
strain C.glutamicum KY10574 at least partially because this strain was actually able to over-
produce L-methionine, got again more influence on production prevention by additional add-
ing of homoserine and homocysteine.

By direct comparison of the progression of the curves for the cultivations with homoserine
and homocysteine the faster production using homocysteine seemed remarkable. But at the
end of the cultivation time both cultivations showed almost the same L-methionine concentra-
tion.

- 99 -
___________________________________________________________________________
4.4.7 Cultivation of Corynebacterium glutamicum KY10574 in the
bioreactor scale (3.5 L)

The bioreactor experiments were performed in bioreactors of the company Infors in a 3.5 L
scale. It was possible to control the aeration rate [0.5 L/(L min)] and the stirring speed in
combination with the oxygen partial pressure (pO
2
). If the pO
2
dropped under 20% the stirring
speed was enhanced automatically to guarantee a better oxygen supply in the cells by higher
mixing rates, which was much more effective than the enhancement of the aeration rate. The
bioreactor cultivation was performed with glucose as the limiting factor to be able to control
growth. This is the state of the art method to cultivate strains of Corynebacterium glutamicum
in the bioreactor scale in order to be able to control cell growth and avoid substrate inhibition
and foaming problems (Kumar et al. 2003; Eggeling et al. 1999). For this reason, the glucose
concentration at the beginning was only around 10 20 g/l, the glucose feed (2 200 ml; 300
g/l) started immediately after the moment when the glucose was consumed with a distinct
pumping rate to guarantee the carbon source supply (fed batch process).

In the following chapter 4 bioreactor cultivations will be shown (in chronological order) to
demonstrate the progress in L-methionine production, the problems (foaming!) and additional
options in the process controlling. Cultivation parameters like aeration rate, pH, foam devel-
opment and glucose feed could be adjusted very precisely in comparison to the shake flasks.
In all cases the the bioreactor cultivations were inoculated with 200 ml of shake flask precul-
tures which were performed for 48 h at 30C with 150 rpm (F1 minimal medium, 100 ml in 1
L shake flasks). So there were new approaches to influence the development of cell growth
and L-methionine production in the bioreactor.

4.4.7.1 Bioreactor cultivation with 10 g/l glucose at t = 0 and a continuous
feeding strategy for additional glucose with high pumping rates

The first bioreactor cultivation with the strain C. glutamicum KY10574 was performed with
an amount of 10 g/l of glucose in the cultivation medium (F1 medium, see Table 3.6). 2.8 L
medium were inoculated with 200 ml of a shake flask preculture. Additionally, a glucose feed
[2 200 ml, 300 g/l, pumping rate mode: 70 ml/h (11 h-14 h), 15 ml/h (since 14 h)] of 400 ml
and 100 ml of anti-foam substances were added to the bioreactor. Altogether 150 g glucose
were given to the cultivation. Figure 4.30 describes the development of the production of the

- 100 -
___________________________________________________________________________
amino acids glycine and L-methionine as well as the development of cell growth (BDM) and
glucose consumption. Only two amino acids, glycine and methionine, were produced in re-
markable amounts, as shown in graph. The biomass increased rapidly to values of 16 g/l after
24 h, which was not reached in any shake flask experiment.

Time (h)
0,0 2,5 5,0 7,5 10,0 12,5 15,0 17,5 20,0 22,5 25,0
B
D
M

(
g
/
l
)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
G
l
u
c
o
s
e

(
g
/
l
)
0
2
4
6
8
10
12
14
16
18
A
m
i
n
o

A
c
i
d

(
m
g
/
l
)
0
100
200
300
400
500
BDM
Glucose
Methionine
Glycine

Start of Feeding
Figure 4.30: Bioreactor cultivation (3.5 L) of C. glutamicum KY10574 over 25 h at 30C in F1 mini-
mal medium with 10 g/l glucose (t = 0); glucose feed (2 200 ml, 300 g/l), pumping rate mode: 70
ml/l (11h-14h), 15 ml/h (since 14h); presented are the courses for the development of the BDM, the
glucose and the produced amino acids L-methionine and glycine; inoculation with 200 ml preculture
(F1 medium) from shake flask cultivation; aeration rate: 0.5 L/(L min); amino acid analysis with GC.

The fast cell growth velocity and biomass accumulation were also a major disadvantage and
problem, because an extreme foaming of the cultivation broth occurred after 24 h which could
not be reduced effectively through the addition of anti-foam reagent (Struktol J673, diluted
1:10 with H
2
O). Foam formation could be detected by a foam sensor which was automatically
connected with pumps for anti-foam reagent. The experiment was finished after 24 h because
the bioreactor began with overfoaming. The cell growth was too fast and not controllable in
this phase of the cultivation. The main reason for this development was the glucose level. As
shown in Figure 4.30, the glucose pool (the initially supplied sugar) was exhausted after 11 h
and the glucose feed was performed afterwards with much too high pumping rates. This led to
the accumulation of the carbon source in the medium and consequently, there was no limita-

- 101 -
___________________________________________________________________________
tion of cell growth any more. The potential of the microorganisms to produce L-methionine
and glycine under these conditions was obvious, within 24 h 425 450 mg/l L-methionine
and 300 mg/l glycine were obtained. These results suggest that even much higher yields of
those amino acids are feasible in cultivations which are performed over longer time periods.

The impacts of glucose accumulation in the medium and the absence of a limitation factor on
the oxygen supply are presented in Figure 4.31. The oxygen partial pressure (pO
2
) dropped
under 20% and the stirring speed was increased automatically to the adjusted limit of 950 rpm
in order to enhance the oxygen supply inside the cultivation broth. Despite the acceleration of
the stirring speed the decrease of the pO
2
values could not be prevented. The increase of the
pO
2
during the last hour (Figure 4.31) occurred even after the cultivation began with over-
foaming. Stirring speed and aeration rate were enhanced in order to avoid the oxygen limita-
tion but they also provoked more foam formation.

Time (h)
0 5 10 15 20 25
S
t
i
r
r
i
n
g

s
p
e
e
d

(
1
/
m
i
n
)
0
200
400
600
800
1000
F
e
e
d

(
-
)
0
10000
20000
30000
40000
p
O
2

(
%
)
0
20
40
60
80
100
A
n
t
i
-
f
o
a
m

(
-
)
0
2000
4000
6000
8000
10000
12000
14000
16000
Stirring speed
Glucose feed
pO
2

Anti-foam

Start of Feeding
mal medium with 10 g/l glucose (t = 0); presented are the courses for the development of the pO
2.
the
stirring speed and the pump rates [70 ml/h (11-14h), 15 ml/h (since 14h)] for the addition of the glu-
cose feed (300 g/l, 2 200 ml) and the anti-foam substance; inoculation with 200 ml preculture (F1
medium) from a shake flask cultivation; aeration rate: 0.5 L/(L min).

- 102 -
___________________________________________________________________________
Figure 4.31 shows also the pumping rates for the addition of the anti-foam reagent and the
glucose (Feed). According to the slow cell growth in the first 11 h of the cultivation (see Fig-
ure 4.30) glucose was not added continuously to the bioreactor but added all at once. The ef-
fects of these crucial occasions on the cell growth and the foam formation had been clearly
underestimated as Figures 4.31 and 4.30 show. The overproportional increase of the addition
of anti-foam reagent at the end of the cultivation could not prevent the strong enhancement of
foam development in the bioreactor.

The exhaust gas analysis is an indicator for the potential of the microorganisms to utilize oxy-
gen from fresh air and to produce carbon dioxide, it is a valuable tool for determination of
metabolic activity. The courses for the development of the oxygen uptake rate (OUR) and the
carbon dioxide evolution rate (CER) are shown in Figure 4.32. This Figure shows also the
respiratory coefficient (RQ). The respiratory coefficent RQ is calculated as the coefficient q
co2

/ q
o2
(referring to molar mass). In Figure 4.32 the connection between the oxygen uptake rate
(OUR) and the carbon dioxide formation rate became clearly obvious. The courses were
nearly parallel; if the OUR course increased, than the CER course developed parallel and rose
too. The potential to transfer oxygen to CO
2
was enhanced if there was more available oxygen
for utilization. The big rise of the courses at the end of the cultivation (maximum at 21,5 h)
can be referred to the exponential cell growth in this time period of cultivation (see Figure
4.30). According to the higher oxygen demand of the microorganisms due to their fast prolif-
eration (uncontrolled growth) and the increased stirring speed the values for OUR and CER
rose overproportionally. The values fell down again after the bioreactor cultivation began
with overfoaming, the outlet for exhaust gas analysis was partially blocked with liquid con-
tents.

The respiratory coefficient RQ shows a course development with values around 1. If the RQ
values are below 1 more carbon dioxide is produced than oxygen is taken from fresh air. Val-
ues above 1 are indicators for the contrary. The course development in the first 5 h (see Figure
4.32) is not meaningful due to very low values for CER and OUR. In general the data from
the exhaust gas analysis (Figure 4.32) have to be seen as a mathematical model which intends
to reflect reality, but not as absolute values like e.g. concentration data for amino acids or the
bio dry mass.

- 103 -
___________________________________________________________________________
Time (h)
0,0 2,5 5,0 7,5 10,0 12,5 15,0 17,5 20,0 22,5 25,0
C
E
R
,

O
U
R

(
m
o
l
/
l
*
h
)
0,00
0,02
0,04
0,06
0,08
0,10
0,12
0,14
0,16
0,18
0,20
CER
OUR

Time (h)
0,0 2,5 5,0 7,5 10,0 12,5 15,0 17,5 20,0 22,5 25,0
R
e
s
p
i
r
a
t
o
r
y

c
o
e
f
f
i
c
i
e
n
t

R
Q

(
-
)
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1,4

Figure 4.32: Course development of the oxygen uptake rate (OUR), the carbon dioxide evolution rate
(CER) and the respiratory coefficient during the cultivation of C. glutamicum KY10574 in the bioreac-
tor (3.5 L) in F1 minimal medium over 25 h at 30C with 10 g/l glucose (t = 0); glucose feed (300 g/l,
2 200 ml), pumping rate mode: 70 ml/h (11 h-14 h), 15 ml/h (since 14 h); inoculation with 200 ml
preculture (F1 medium) from shake flask cultivation; aeration rate: 0.5 L/(L min).

- 104 -
___________________________________________________________________________
feeding strategy for additional glucose with low pumping rates

This bioreactor cultivation was performed under consideration of the results formerly ob-
tained in the cultivation which was described in the chapter 4.4.7.1. The glucose concentra-
tion at the beginning was 10 g/l, the glucose feed was performed in difference to the first bio-
reactor cultivation with very low pumping rates (10 ml/h, 2 200 ml, 300 g/l). As shown in
Figure 4.33, on the one hand, further controllable cell growth could be guaranteed and on the
other hand glucose accumulation could be avoided. During the continuous feeding process the
glucose concentration was located always nearly zero, therefore, controlled growth under
limitation conditions was established. The bioreactor medium (2.8 L) was inoculated with 200
ml from shake flask preculture (48 h). In addition 400 ml glucose feed and 100 ml anti foam
substances were given in the bioreactor during the cultivation time. During cultivation 150 g
glucose were given altogether to the medium. In comparison to the first bioreactor cultivation
(Figure 4.30) the cell growth slowed down; 11.5 g/l BDM were generated after 48 h. Amounts
like these and more were achieved in the first cultivation already after 24 h but the bioreactor
tended to foam over if the cell growth was too fast. Although the bio dry mass even after 48 h
was lower than in the first cultivation after 24 h, these conditions are more favourable for the
process because the problems related to foam began at a much later point in time (42 h, see
Figure 4.34). The L-methionine and glycine concentrations increased continuously to values
of 920 mg/l (methionine) and 500 mg/l (glycine) after 48 h towards the end of the cultivation.

After 48 h the cultivation was finished due to intense foaming and the overproportional addi-
tion of anti foam substance (see Figure 4.34). For this reason the amino acid production
course stopped in maximum. In further cultivations over longer time periods it might be pos-
sible to obtain even higher concentrations of the amino acids methionine and glycine. But as a
prerequisite, an effective process to reduce the foam has to be worked out.

- 105 -
___________________________________________________________________________

Time (h)
0 10 20 30 40 50
B
D
M

(
g
/
l
)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
G
l
u
c
o
s
e

(
g
/
l
)
0
2
4
6
8
10
12
A
m
i
n
o

A
c
i
d

(
m
g
/
l
)
0
100
200
300
400
500
600
700
800
900
1000
BDM
Glucose
Methionine
Glycine

Feed 2 on
Feed 1 on
mal medium with 10 g/l glucose (t = 0); glucose feed with a pumping rate of 10 ml/h (2 200 ml, 300
g/l), presented are the courses for the development of the biomass, the glucose and the produced amino
acids L-methionine and glycine; inoculation with 200 ml preculture (F1 medium) from shake flask
cultivation; aeration rate: 0.5 L/(L min); amino acid analysis with GC.

It can be quite clearly stated that the glucose limitation factor after 12 h was the crucial culti-
vation, the low pumping rates for glucose feed addition made cell growth and product devel-
opment more controllable although the foaming at the end of the cultivation was very strong.
In contrast to the former experiment (Figure 4.30) with higher accumulated glucose concen-
trations and higher feeding rates (Figure 4.31) during the cultivation time, this time foaming
could be reduced significantly. Not only that bioreactor cultivation is more controllable with
one parameter responsible for limited growth, also the amino acid concentrations would be
much more higher by distributing the glucose addition over longer time periods.

The development of pO
2
, the stirring speed and the pumping rates for the addition of glucose
and anti foam substance are shown in Figure 4.34. The pO
2
did not drop under 20% during
the cultivation by a constant stirring speed of 700 rpm; so there were no phases of extreme
anaerobic conditions with diminished cell growth and production in the bioreactor as it oc-

- 106 -
___________________________________________________________________________
curred in former experiments (Figure 4.31). In comparison to experiments in the shake flasks
there were no more remarkable amino acid side products like alanine or valine, this indicated
that the production conditions for L-methionine and glycine were especially well adapted in
the bioreactor. The strong foam formation (see Figure 4.34) towards the end of the cultivation
could be evidence for another limitation by another nutrition source. However, the biomass
growth curve was still increasing in this time period (Figure 4.33).

Time (h)
0 10 20 30 40 50
S
t
i
r
r
i
n
g

s
p
e
e
d

(
1
/
m
i
n
)
0
200
400
600
800
1000
F
e
e
d

(
-
)
0
10000
20000
30000
40000
A
n
t
i
-
f
o
a
m

a
d
d
i
t
i
o
n

(
-
)
0
20000
40000
60000
80000
p
O
2

(
%
)
0
20
40
60
80
100 Stirring speed
Glucose feed
Anti-foam
pO
2

Feed 2 on Feed 1 on
2.
the
stirring speed and the pump rates for the addition of the glucose feed (10 ml/h, 300 g/l, 2 200 ml) and
the anti-foam reagent (Struktol J673); inoculation with 200 ml preculture (F1 medium) from shake
flask cultivation; aeration rate: 0.5 L/(L min).

In Figure 4.35 the oxygen uptake rate (OUR) and the carbon dioxide formation rate (CER) are
shown. The progressions of both curves interdepend strongly on each other. The big rise
(maximum at 18h) and the big descent (minimum at 36 h) of the curve courses are referring to
glucose limitation or absence (compare with Figure 4.32), the rise fits perfectly to the expo-
nential cell growth phase (see Figure 4.33). Generally, the fluctuations of carbon dioxide for-
mation (CER) were slightly higher. This parameter was more sensitive than the oxygen up-
take rate (OUR) in the experiments and reacted very quickly to variations in the metabolism.

- 107 -
___________________________________________________________________________
Time (h)
0 5 10 15 20 25 30 35 40 45 50
C
E
R
,

O
U
R

(
m
o
l
/
l
*
h
)
0,00
0,01
0,02
0,03
0,04
0,05
CER
OUR

Time (h)
0 10 20 30 40 50
R
e
s
p
i
r
a
t
o
r
y

c
o
e
f
f
i
c
i
e
n
t

R
Q

(
-
)
0,0
0,5
1,0
1,5
2,0
2,5

(CER) and the respiratory coefficient during the cultivation of C. glutamicum KY10574 in the bioreac-
tor (3.5 L) in F1 minimal medium over 48 h at 30C with 10 g/l glucose (t = 0); glucose feed (10 ml/h,
300 g/l, 2 200 ml); inoculation with 200 ml preculture (F1 medium) from shake flask cultivation;
aeration rate: 0.5 L/(L min).

- 108 -
___________________________________________________________________________
The respiratory coefficient RQ (Figure 4.35) values were most of the time slightly above 1,
with the exceptions of two big peaks after 18 h and 34 h. These major fluctuations can be re-
ferred to nutrient absence factors (glucose, see Figure 4.33), which are responsible for the cell
metabolism acceleration or decceleration. In general, during the bioreactor cultivation more
carbon dioxide was produced than oxygen was consumed. One reason for this property is the
fact that the substrate glucose (C
6
H
12
O
6
) is very rich in oxygen which can be used for meta-
bolic reactions. Therefore, it was not necessary for the microorganisms to consume higher
amounts of oxygen from the fresh air.

feeding strategy for additional glucose with low pumping rates

After the first experiences with bioreactor cultivations of Corynebacterium glutamicum
KY10574 in further experiments it was intended to enhance both the cell growth velocity and
the L-methionine production. As a first step the glucose concentration in the start medium
was increased to 20 g/l to improve and accelerate the cell growth of the microorganisms be-
fore starting the glucose feeding (Figure 4.36). A compromise between enhanced but control-
lable cell growth and a low occurrence of foam formation had to be found.

The bio dry mass course (Figure 4.36) increased much faster than in the comparable time
ranges in former cultivations (see Figure 4.33). At start of the glucose feed 1, there was al-
ready a biomass concentration of 8 g/l; in the former cultivations with 10 g/l glucose initially
supplied to the medium (see Figures 4.33 and 4.30) approximately 2.2 g/l were reached at this
point in time. Thus, the doubling of glucose concentration right at the beginning elevated the
BDM overproportionally. At the end of the cultivation process nearly 14 g/l biomass were
produced. L-methionine production could also be enhanced to values of 1.15 g/l, approxi-
mately 200 mg/l more than in the former experiment (chapter 4.4.7.2). L-glycine was gained
in almost the same amount of 500 mg/l after 48 h. The L-methionine production phase began
with almost no delay very rapidly after the start of the cultivation. In the time period between
12 h and 30 h the L-methionine concentration increased quickly from 0.3 g/l to 1.1 g/l, after-
wards the concentration remained almost constant during the last 18 h of the cultivation time.

- 109 -
___________________________________________________________________________

Time (h)
0 10 20 30 40 50
G
l
u
c
o
s
e

(
g
/
l
)
0
10
20
30
40
O
D
5
4
6
0
20
40
60
80
B
D
M

(
g
/
l
)
0
2
4
6
8
10
12
14
16
18
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
200
400
600
800
1000
1200 Glucose
BDM
OD
546
Methionine
Glycine

Feed 2 on Feed 1 on
mal medium with 20 g/l glucose (t =0); glucose feed with a pumping rate of 10 ml/h (2 200 ml, 300
g/l), presented are the courses for the development of the BDM, the glucose and the produced amino
acids L-methionine and glycine; inoculation with 200 ml preculture (F1 medium) from shake flask
cultivation; aeration rate: 0.5 L/(L min); amino acid analysis with GC.

In spite of all efforts to avoid glucose accumulation a slight increase in the concentration after
36 h was observed. This amount was again consumed until the end of the process. The reason
for the accumulation could be in pumping rates which might have been adjusted too highly, or
in the inhibition of the cell growth through nutrient limitation. The biomass was already in the
stationary phase at this time, which was an additional indicator for limitation conditions. It
might be concluded that the microorganisms reorganized their metabolic pathways at that
time in order to use alternative biochemical precursors and additional biochemical pathways
in order to maintain a certain level of cell population. This is the current state of the art
method (Eggeling et al. 1999; Gomes et al. 2005) to provide a balanced metabolism caused by
nutrient shortage, at least for short-term periods.

A visual presentation of the development of the L-methionine values is shown in Figure 4.37.
The development of the increased L-methionine concentration can be seen by comparing the

- 110 -
___________________________________________________________________________
thickness of the spots and in relation to the 0.5 g/l L-methionine reference, but it needs to be
emphasized that this method is only useful for qualitative detections.

Meth.
0.5 g/l
24.2 h 45.4 h 36.2 h 12.2 h 17.5 h 6.3 h

Figure 4.37: Qualitative thin layer chromatography (TLC) of the samples obtained from the bioreactor
cultivation (Figure 4.36) and a reference of L-methionine (500 mg/l); detection with ninhydrin reac-
tion.

Foam formation could not be avoided completely, particularly not during the last 18 h of the
cultivation process (see Figure 4.38). But the foam did not tend to expand too much; on the
contrary, there was a relatively stabilised amount of foam on top of the liquid surface. For this
reason the addition of anti-foam reagent was ceased after 36 h. The foam formation in the last
phase of the cultivation time was most probably caused through collapsing cells (nutrient de-
ficiency) and the released contents from the cytosol, e.g. proteins, which can favour foaming
extensively (Chmiel 2006). The pO
2
sensor system did not work correctly during this cultiva-
tion. Therefore, the stirring speed was constantly adjusted at 600 rpm (0 12 h) at the begin-
ning and at 700 rpm (12 h 48h) afterwards in order to be aware of changes in oxygen de-
mand with increased cell density.

- 111 -
___________________________________________________________________________
Time (h)
0 1

0 20 30 40
S
t
i
r
r
i
n
g

s
p
e
e
d

(
1
/
m
i
n
)
0
200
400
600
800
1000
p
H

0
2
4
6
8
10
A
n
t
i
-
f
o
a
m

a
d
d
i
t
i
o
n

(
-
)
0
10000
20000
30000
40000
50000
60000
70000
F
e
e
d

(
-
)
0
20000
40000
60000
Feed 1 on Feed 2 on
Stirring speed
pH
Anti-foam
Glucose feed

mal medium with 20 g/l glucose (t = 0); presented are the courses for the development of the pH, the
stirring speed and the pumping rates for the addition of the glucose feed (10 ml/h, 300 g/l, 2 200 ml)
and the anti-foam reagent (Struktol J673); inoculation with 200 ml preculture (F1 medium) from shake
flask cultivation; aeration rate: 0.5 L/(L min).

The development of the pH is also shown in Figure 4.38. This parameter was measured
online, but there was no regulation or controlling of its progression. The values ranged be-
tween 8 and 5.5 at the end of the cultivation time.

Figures 4.39 presents the exhaust gas analysis of this bioreactor cultivation. The movements
in the courses for oxygen uptake rate (OUR) and carbon dioxide evolution rate (CER) mostly
refer to the short term complete absence of glucose and the connected alterations in cell
growth velocity which was removed quickly by starting the feeding process (Feed 1 and 2, see
Figures 4.38 and 4.36). The respiratory coefficient (RQ) course development showed values
around 1 with the exception of the last 10 hours of the cultivation when the values fell down
under 0.5. At this point in time the demand for oxygen and consequently the formation of
carbon dioxide decreased rapidly due to restricted growth conditions (glucose deficiency).

- 112 -
___________________________________________________________________________
Time (h)
0 10 20 30 40 50
C
E
R
,

O
U
R

(
m
o
l
/
l
*
h
)
0,00
0,01
0,02
0,03
0,04
0,05
CER
OUR

Time (h)
0 10 20 30 40 50
R
e
s
p
i
r
a
t
o
r
y

c
o
e
f
f
i
c
i
e
n
t

R
Q

(
-
)
0,0
0,5
1,0
1,5
2,0
2,5

(CER) and the respiratory coefficient (RQ) during the cultivation of C. glutamicum KY10574 in the
bioreactor (3.5 L) over 48 h at 30C in F1 minimal medium with 20 g/l glucose (t = 0); glucose feed
(10 ml/h, 300 g/l, 2 200 ml); inoculation with 200 ml preculture (F1 medium) from shake flask culti-
vation; aeration rate: 0.5 L/(L min).

- 113 -
___________________________________________________________________________
feeding strategy for additional glucose with low / mid pumping rates
and an effective pO
2
control system

This cultivation was performed because the pO
2
sensor did not work properly in the former
experiment described in the chapter 4.4.7.3. Therefore, it was not possible to observe the im-
pact of the automatic control between the stirring speed and the partial oxygen pressure (pO
2
)
on cell growth and amino acid production. The automatic control began to increase the stir-
ring speed as soon as the pO
2
dropped under 30%. This should guarantee an optimal oxygen
supply inside the bioreactor and high cell growth velocities. In former investigations (shake
flask and bioreactor, see e.g. chapter 4.4.3.7) it was clearly obvious that high product concen-
trations of L-methionine were only available if the oxygen supply was sufficient, which was a
crucial parameter in all experiments and investigations.

Time (h)
0 10 20 30 40 50
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
0
200
400
600
800
1000
1200
1400
1600
G
l
u
c
o
s
e

(
g
/
l
)
0
10
20
30
40
O
D
5
4
6
0
20
40
60
80
100
B
D
M

(
g
/
l
)
0
5
10
15
20
Methionine
Glucose
OD
546
BDM
Glycine

Feed 1 on Feed 2 on
mal medium with 20 g/l glucose (t = 0), glucose feed with a pumping rate of 10 - 15 ml/h (2 200 ml,
300 g/l), presented are the courses for the development of the BDM, the glucose and the produced
amino acids L-methionine and glycine; inoculation with 200 ml preculture (F1 medium) from shake
flask cultivation; aeration rate: 0.5 L/(L min); amino acid analysis with GC.

- 114 -
___________________________________________________________________________
As shown in Figure 4.40, the carbon source glucose was exhausted after 13 hours. The
amount of the bio dry mass even increased to nearly 16 g/l after 48 h compared to the former
cultivation with 14 g/l of bio dry mass in maximum and 20 g/l of glucose at the beginning
(see Figure 4.36). The optical density (OD
546
) reached values around 70. The L-methionine
concentration values still increased when the measured values for the optical density and the
BDM indicated the stationary cell growth phase. The maximum L-methionine value of 1.45
g/l was achieved after 44 h, afterwards it dropped slightly at the end of the cultivation period.
A more visual and qualitative view of the development of the L-methionine concentrations
during the cultivation is presented with the thin layer chromatography (TLC) in Figure 4.41.
The glycine amount could be enhanced to values higher than 650 mg/l, this is more than dou-
bled compared with the maximum in the first bioreactor cultivation (300 mg/l, Figure 4.30). It
can be concluded that the glycine concentrations increase parallel to the L-methionine con-
centrations. Therefore, higher oxygen ratios in the cultivation broth favour also the metabolic
activity of the microorganisms to produce glycine.

6.3 h
Meth.
0.5 g/l
12.3 h 24.3 h 35.6 h 41.3 h 48.0 h

Figure 4.41: Qualitative thin layer chromatography (TLC) of the samples obtained from the bioreactor
cultivation (Figure 4.40) and a reference of L-methionine (500 mg/l); detection with ninhydrin.
- 115 -
___________________________________________________________________________
The major problem in this cultivation was again the foaming. Figure 4.42 shows clearly that a
large amount of foam was produced during the last 12 h of the cultivation. Therefore, 200 ml
anti-foam reagent was pumped into the bioreactor to reduce foam formation, which was twice
the amount compared to the other bioreactor experiments. The connection between the curves
for the oxygen partial pressure (pO
2
) and the stirring speed is a well established fact (Chmiel
2006). The stirring speed increased immediately when the pO
2
values decreased below 30%.
On the other hand the stirring speed would decrease as soon as the values for the pO
2
were
above 30%. This automatic control mechanism provided optimised oxygen supply and subse-
quently elevated cell growth. In this case cell growth was enhanced in such a manner that the
production of L-methionine was increased to values around 1.45 g/l (see Figure 4.40). But as
a consequence, the high stirring speeds and the massive cell mass accumulation caused also
intensive and enormous foaming. The process performance became even more difficult.
Therefore, for further experiments a compromise needs to be found between high product
accumulation on the one hand and lower cell growth and foaming on the other hand.

Time (h)
0 10 20 30 40
S
t
i
r
r
i
n
g

s
p
e
e
d

(
1
/
m
i
n
)
0
200
400
600
800
1000
A
n
t
i
-
f
o
a
m

a
d
d
i
t
i
o
n

(
-
)
0
20000
40000
60000
80000
100000
F
e
e
d

(
-
)
0
10000
20000
30000
40000
p
O
2

(
%
)
0
20
40
60
80
100
Stirring speed
Anti-foam add.
Glucose feed
pO
2

Feed 1 on Feed 2 on
2
, the
stirring speed and the pump rates for the addition of the glucose feed (10 - 15 ml/h, 300 g/l, 2 200
ml) and the anti-foam reagent (Struktol J673, 200 ml); inoculation with 200 ml preculture (F1 me-
dium) from shake flask cultivation; aeration rate: 0.5 L/(L min).

- 116 -
___________________________________________________________________________
The exhaust gas analysis results are shown in Figure 4.43. There was a signal recording prob-
lem after 43 h; therefore, data for the last hours were not available. The rise of the CER and
OUR courses at the beginning (maximum at 12 h) fit perfectly with the exponential biomass
accumulation (see Figure 4.40). It can be concluded by comparing these data with the infor-
mation shown in the Figure 4.42 + 4.40 that the metabolic activities of the microorganisms
dropped significantly after 40 h. The cells began to collapse and to release a broad variety of
contents from the cytosol, some of them were well known as foam causing substances (e.g.
proteins). The reasons for the strong foam formation between 25 h and 30 h were most proba-
bly the high cell growth rates and the high stirring speed rates; these attendant circumstances
favoured the foam formation. A method to inhibit the extensive foaming and to provide even
improved conditions for oxygen supply might be the gassing of the cultivation broth with pure
oxygen, although handling with pure oxygen is not easy. A further possibility to avoid foam-
ing is the application of different chemical anti-foam reagents like e.g. sun flower oil or rape-
seed oil (colza oil).

The development of the respiratory coefficient (RQ) in Figure 4.43 was characteristic for all
cultivations in the bioreactor with a course development around 1. Although the oxygen sup-
ply conditions were improved for this cultivation compared to the last one, these alterations
caused simultaneously increased cell growth and even higher oxygen demands. The im-
provement of the oxygen supply increased automatically the cell mass formation and subse-
quently led to even higher demands for oxygen. The production of carbon dioxide during the
cultivation of C.glutamicum KY10574 cells was higher than the uptake rate of oxygen from
fresh air due to the fact, that the microorganisms are able to use the oxygen parts of the sub-
strate glucose (C
6
H
12
O
6
) for metabolism activities.

- 117 -
___________________________________________________________________________
Time (h)
0 10 20 30 40 50
C
E
R
,

O
U
R

(
m
o
l
/
l
*
h
)
0,00
0,02
0,04
0,06
0,08
CER
OUR

Time (h)
0 10 20 30 40 50
R
e
s
p
i
r
a
t
o
r
y

c
o
e
f
f
i
c
i
e
n
t

R
Q

(
-
)
0,0
0,5
1,0
1,5
2,0
2,5

(CER) and the respiratory coefficient (RQ) during the cultivation of C. glutamicum KY10574 in the
bioreactor (3.6 L) in F1 minimal medium over 48 h at 30C with 20 g/l glucose (t = 0); glucose feed
(10 - 15 ml/h, 300 g/l, 2 200 ml); inoculation with 200 ml preculture (F1 medium) from shake flask
cultivation; aeration rate: 0.5 L/(L min).

- 118 -
___________________________________________________________________________
4.4.7.5 Summary about the bioreactor cultivations with Corynebacterium
glutamicum KY10574

Considering the results from the bioreactor experiments the strategy to perform the cultiva-
tions with a low initially supplied glucose concentration and a following continuous feeding
strategy was successful. Additionally, this was absolutely necessary in order to avoid massive
foam formation. It was important to get control over cell growth velocity and foam formation
using a limitation factor as glucose to be able to cultivate the microbial strain under best con-
ditions for amino acid production and process handling. Anaerobic conditions in the cultiva-
tion broth resulted directly in a significantly diminished production of L-methionine (see Fig-
ure 4.30).

It was advantageous to double the initially supplied sugar concentration from 10 to 20 g/l in
order to enhance the cell growth velocity and subsequently the L-methionine production in
the first hours of the cultivation until the glucose was consumed. Usually, the foam formation
appeared in phases of nutrient deficiency (glucose) and under anaerobic conditions. The an-
aerobic conditions occurred when the cell growth was too fast (no effective limitation through
the sugar concentration) and the consumption of oxygen was too strong (see Figure 4.31).
Proteins and other foam developers were released after the cell collapse. Therefore, the high-
est L-methionine concentrations [1.15 g/l (Figure 4.36) and 1.46 g/l (Figure 4.40)] could be
achieved using 20 g/l of initially supplied sugar and an effective mechanism for oxygen trans-
port into the liquid cultivation broth by automatic control between the partial oxygen pressure
and the stirring speed (see Figure 4.42).

Table 4.2 shows the development of significant parameters in comparison of all bioreactor
cultivations. Apart from the total amount of glucose, the used pumping rates for the glucose
feed and the maximum reached values for the bio dry mass and L-methionine concentrations,
there is an overview about the development of the yield coefficients (Y
X/S
, Y
P/S
, Y
P/X
). The
values for these coefficients are applied to the data derived at the end of the cultivations. It
can be seen in Table 4.2 that the value for the yield coefficient Y
X/S
from bioreactor cultiva-
tion 1 (chapter 4.4.7.1) with 0.37 was the highest one. The conversion from substrate to bio-
mass was most effective in this cultivation; the maximum bio dry mass value in this cultiva-
tion (16.2 g/l, see Table 4.2) was also the highest compared to the other cultivations. The val-
ues for the yield coefficient Y
X/S
drop to a minimum of 0.21 (bioreactor cultivation 3) and rise
again to 0.3 at the end of bioreactor cultivation 4.

- 119 -
___________________________________________________________________________
On the other side the values for the yield coefficient Y
P/X
show an increase from 0.025 (first
cultivation) to 0.094 in the fourth cultivation. The ratio of product (P, L-methionine) to bio
dry mass (X) is much more advantageous in the cultivations with an effective limitation of the
cell growth through glucose limitation. The same trend is visible for the yield coefficient Y
P/S
;
the values for this coefficient increased continuously from the first (0.095) to the fourth culti-
vation (0.028). So the product oriented yield coefficients (Y
P/S
, Y
P/X
) emphasize the impor-
tance of a continuous fed batch cultivation process.

Therefore, the yield coefficients reveal that the first bioreactor cultivation was effective in fast
biomass accumulation but very non-effective in the L-methionine production which was the
crucial factor in all experiments. On the contrary, it is shown that too high cell growth veloci-
ties which caused anaerobic conditions in the cultivation broth are extremely inapprobriate for
the L-amino acid production (see the course development for the oxygen partial pressure pO
2

in Figure 4.31 and corresponding values for L-methionine production in Figure 4.30).

Table 4.2: Comparison of the 4 bioreactor cultivations (BR) which are described in the chapter 4.4.7
on the basis of significant biochemical engineering parameters, presented in chronological order from
chapter 4.4.7.1 4.4.7.4 (BR 1-4); the values for yield coefficients (Y
X/S
/ Y
P/S
/ Y
P/X
) are applied to
the data obtained at the end of the cultivations.

Bioreactor (BR) and ini-
tially supplied glucose
BR 1: 10 g/l BR 2: 10 g/l BR 3: 20 g/l BR 4: 20 g/l
Pumping rate mode for
glucose feed (ml/h)
70 ml/h (11-14
h); 15 ml/h
10 ml/h 10 ml/h 10 15 ml/h
Glucose total (g) 150 150 180 180
BDM
max
(g/l) 16.2 11.5 12.9 15.6
L-Methionine
max
(mg/l) 442 934 1,154 1,467
Y
X/S
(g bio dry mass /
g glucose)
0.37 0.27 0.21 0.30
Y
P/X
(g L-methionine /
g bio dry mass)
0.025 0.081 0.092 0.094
Y
P/S
(g L-methionine /
g glucose)
0.010 0.021 0.022 0.028
Glycine
max
(mg/l) 305 511 539 669

- 120 -
___________________________________________________________________________
Table 4.2 shows also the development of the glycine concentrations. It is remarkable that gly-
cine is, apart from L-methionine, the only amino acid which was produced in the bioreactor
cultivations in noticeable amounts. Glycine was also a product in the shake flask cultivations
(see e.g. Figures 4.19 -4.24). The glycine concentrations also increased considerably and in
parallel to the increasing L-methionine concentrations (compare Figures 4.30, 4.33, 4.36,
4.40) but were always clearly below the L-methionine concentrations. In the first bioreactor
cultivation (chapter 4.4.7.1) with partially anaerobic conditions and unsufficient oxygen sup-
ply the glycine values were around 300 mg/l in maximum, in the last cultivation with im-
proved oxygen conditions the maximum value was 669 mg/l; so the concentration was more
than doubled. Therefore, it can be concluded that the glycine production is favoured in the
same way as L-methionine production by improved oxygen supply and oxygen transport into
the cultivation broth.

The amino acids L-alanine, L-valine and in some cases L-glutamic acid which were produced
in noticeable amounts in the shake flask cultures (see Figures 4.19 4.24) disappear in the
bioreactor cultivations. The better conditions for oxygen supply in the bioreactor scale favour
the production of L-methionine and glycine; the production of other amino acid by-products
stopped completely.

In the case of L-alanine, which was in some cases the dominant produced amino acid in the
shake flasks (Figures 4.22 + 4.23), there is another explanation. There was no production due
to the glucose limited conditions in the bioreactor cultivations. As shown before in the shake
flask cultivations (Figures 4.19 4.24) the L-alanine concentrations dropped rapidly after the
glucose was consumed or occurred only in low concentrations in the media.

The bioreactor experiments which are described in this thesis are fed batch cultivations
(closed systems). In following experiments in the bioreactor scale it would be interesting to
observe the influence of a continuous cultivation process in bioreactor with a continuous feed-
ing of nutrient solutions and a continuous draining off (open system) on the L-methionine
accumulation in the cultivation broth. This process will be performed usually in continuous
stirred tank reactor (CSTR) systems.

- 121 -
___________________________________________________________________________
4.4.8 Investigation of total L-methionine concentrations (extra- and
intracellular) of the strain C. glutamicum KY10574 compared
to the strain C. glutamicum DSM20300

C. glutamicum is able to enrich distinct amounts of amino acids inside the cells, as has been
investigated e.g. for glutamic acid production (Clement et al. 1986). Several experiments were
carried out to improve the release of amino acids from the intracellular volume to the extracel-
lular milieu, e.g. by modification of the cell wall density through biotine limitation during
cultivation (Leuchtenberger 1996). The release of the amino acid molecules is important in
order to prevent the deactivation of amino acid production because the cells are constrained to
end this process in order to protect their stability which is at risk through higher enrichments
of amino acids in the intracellular milieu.

0
100
200
300
400
500
600
700
800
900
1000
1100
1200
C.glut. DSM20300 C.glut. KY10574 (cell
growth in declining
phase)
C.glut. KY10574
(stationary cell growth
phase)
M
e
t
h
i
o
n
i
n
e

i
n
t
r
a
c
e
l
l
.

(
m
g
/
k
g

B
D
M
)
0
200
400
600
800
1000
1200
M
e
t
h
i
o
n
i
n
e

e
x
t
r
a
c
e
l
l
.

(
m
g
/
l
)
Methionine intracellular (mg/kg BDM)
Methionine extracellular (mg/l)

Figure 4.44: Comparison of C. glutamicum DSM20300 and KY10574 intracellular amounts of L-
methionine, measured after cell disruption with an ultrasonic homogenizer, the extracellular amounts
are shown for comparison.

The measured amounts of L-methionine (Figure 4.44) after cell disruption were for both
strains extremely minimal, values around 420 mg/kg bio dry mass (C. glutamicum
DSM20300) and 643.9 / 1124.9 mg/kg bio dry mass (C. glutamicum KY10574) could not be

- 122 -
___________________________________________________________________________
considered as an intracellular enrichment of L-methionine. The strain C. glutamicum
DSM20300 which was not able to produce L-methionine in remarkable amounts (in this case
3.7 mg/l) showed almost the same intracellular amount of L-methionine as the strain C. glu-
tamicum KY10574. It is normal for the strain C. glutamicum KY10574 to exhibit more L-
methionine intracellularly in the stationary phase than in the decline phase, because in this
phase the metabolism is still partially working and producing L-methionine which has to be
found naturally in the cytosol before releasing it to the outside. Under these circumstances
even higher amounts would have been expected by measuring of L-methionine in the expo-
nentiell production phase of the strain. It was even astonishing that the results for L-
methionine concentrations in the cytosol were that low compared to the high amounts of L-
methionine measured in the supernatant (see Figure 4.44). The results obtained from these
experiments clearly point out that the strain Corynebacterium glutamicum KY10574 has no
problems releasing its amino acids in the extracellular volume by cultivating in F1 medium.

4.4.8.1 Protein hydrolysis of C. glutamicum KY10574 bio dry mass

The complete protein hydrolysis of C. glutamicum KY10574 was performed to get a result
about the complete L-methionine content obtained in the bioreactor cultivation which is de-
scribed in chapter 4.4.7.3. At the end of the cultivation 12.45 g/l bio dry mass was produced;
the concentration of L-methionine in the supernatant was 1151.15 mg/l (see Figure 4.36). The
bio mass was separated from the supernatant and freeze dried (after a washing procedure).
After total hydrolysis of the biomass including proteinsa determination of the amino acid de-
gree of methionine, threonine and lysine in the bio dry mas was performed. These amino ac-
ids are the most important additives in animal feed nutrition. The hydrolysis and the amino
acid determination was carried out at the LUFA (Landwirtschaftliche Untersuchungs- und
Forschungsanstalt) in Speyer. The results are shown in Table 4.3.

Table 4.3: Amino acid content of the bio dry mass from C. glutamicum KY10574 after protein hy-
drolysis, obtained at the end of the bioreactor cultivation (chapter 4.4.7.3).

Methionine (%) Lysine (%) Threonine (%)
1,47 4,33 3,52

- 123 -
___________________________________________________________________________
The results shown in Table 4.3 had to be converted in order to combine them with the me-
thionine amount in the supernatant. At the end of the bioreactor cultivation 12.45 g/l bio dry
mass (BDM) were gained. After connecting these values with the rule of proportion method,
the results of amino acids obtained after protein hydrolysis can be specified as concentrations,
as shown in Figure 4.45. It was pointed out clearly, that in all shake flask and bioreactor ex-
periments with this strain the major amount of methionine is referred to the supernatant but
for lysine and threonine amounts the major source was the bio mass content.

0
200
400
600
800
1000
1200
1400
1600
methionine lysine threonine
A
m
i
n
o

a
c
i
d

(
m
g
/
l
)
supernatant
protein hydrolysis
supernatant +protein hydrolysis

Figure 4.45: Complete contents of the amino acids methionine, lysine and threonine obtained in bio-
reactor cultivation (chapter 4.4.7.3, see Figure 4.36); presented are the concentrations in the super-
natant, the bio dry mass (BDM) and both summed up.

- 124 -
___________________________________________________________________________
As for L-lysine, for the industrial application the complete cultivation broth was dried (e.g.
spray drying) and mixed to animal feed in suitable ratios (see Figure 2.8). This procedure
avoids the additional costs for down stream processing and makes the use of this biotechni-
cally gained amino acid as feed additive more competitive against the chemical process
(Leuchtenberger 1996). It has been recommended to supply laying hens with animal feed con-
taining approximately 0.3% DL-methionine and 0.61% L-lysine (Keshavarz 2003; Noftsger et
al. 2003; Misciatelli et al. 2003). The supplementation with lysine, methionine and threonine
is necessary because the established sources for animal feed (wheat, soy bean) exhibit defi-
ciencies of these amino acids compared to the nutritional requirements of poultry breeding
(Pack 2004).

- 125 -
Summary
___________________________________________________________________________
5 Summary / Outlook

The first objective at the beginning of this research was the screening for L-methionine over-
producers using Corynebacterium glutamicum DSM20300. This strain has the capability to
produce L-glutamic acid and was the starting point for numerous studies on amino acid pro-
duction. The method to generate the overproducing mutants was the random mutagenesis by
ultraviolet radiation and following selection with methionine analogues / anti-metabolites, e.g.
ethionine or norleucine. It was possible to create mutants and to measure production capabili-
ties under high throughput conditions. The selection system turned out to be the crucial prob-
lem. The application of anti-metabolites was a necessary, but not a sufficient criterion. The
microorganisms which probably owned the abilities to overproduce L-methionine had to grow
on media with anti-metabolite contents. It was not imperative, however, that these mutants
gained the properties to produce L-methionine. In spite of all experimental efforts, an over-
producer of L-methionine could not be obtained by this method.

The strain Corynebacterium glutamicum ATCC21608 was bought from the ATCC because
indications on L-methionine production abilities had been reported (Nakayama et al. 1973).
After several tests in shake flask cultivations with different media it could be proved that the
delivered strain material was able to overproduce L-alanine (ca. 200 mg/l), but not L-
methionine. Perhaps the strain material was too old and inactive after a long period of storage
at the ATCC.

The Japanese company Kyowa Hakko Kogyo Ltd. kindly provided the strain Corynebacte-
rium glutamicum KY10574. This strain was able to produce L-methionine.

In the first experiments with this strain in the shake flask scale in a combination of a complex
preculture and a minimal main medium a broad variety of amino acids (e.g. lysine, glycine,
alanine, methionine) could be observed. Lysine was the dominant amino acid (nearly 700
mg/l) under those conditions, methionine was produced in amounts of ca. 125 mg/l. The pro-
duction of lysine began in the phase of precultivation in the complex medium, therefore, the
product spectrum was preadjusted to lysine through the usage of complex ingredients.

- 126 -
Summary
___________________________________________________________________________
The use of a minimal medium which was able to deliver all nutrients for the cell growth of
Corynebacterium glutamicum KY10574 provided a way to alter the product spectrum, so that
a complex precultivation was not necessary any more. It was possible to establish a minimal
medium which guaranteed best cell growth conditions. The product spectrum was changed
completely, there was no remarkable production of lysine any more. L-methionine was now
the dominant amino acid, followed by glycine. It could be shown in several shake flask ex-
periments with this medium that high shaking rates and a distinct concentration of glucose
(around 55 g/l) could improve the production capability of L-methionine considerably. Suffi-
cient oxygen supply was important for rising the L-methionine production with C. glu-
tamicum, and values of 1.45 g/l L-methionine could be achieved using a preculture. The use
of higher glucose concentrations than 70 g/l favoured the production of another amino acid,
L-alanine.

In the bioreactor scale, it was necessary to cultivate Corynebacterium glutamicum KY10574
with a limitation factor in order to control cell growth velocity and to guarantee sufficient
oxygen supply for optimised L-methionine production. The glucose concentration was used in
the cultivations as limitation factor, the balance between high cell growth and L-methionine
production on the one hand and controlled growth with sufficient oxygen rates on the other
hand could be improved crucially. The production of L-methionine in the bioreactor scale
could be enhanced from 425 mg/l in the first cultivation to 1.4 g/l in the last cultivation. A
major problem was foam formation in the bioreactor due to uncontrolled growth with inten-
sive cell mass accumulation and the cell collapse under anaerobic and nutrient limited condi-
tions at the end of the cultivations. This problem could be solved mostly by an automatic
foam sensor system connected with an anti-foam reagent feed. But still high amounts of this
reagent were necessary to prevent foaming in an adequate way.

The major differences in amino acid production after media alteration emphasized the impor-
tance of the medium components for L-methionine production. After extensive investigations
on the influence of the carbon source (glucose) on the L-methionine production, the im-
provement and optimisation of the nutrients providing ammonia (NH
4
+
), sulphate (SO
4
2-
) and
phosphate (PO
4
3-
) could also help to enhance production. With L-methionine overproducing
strain C. glutamicum KY10574 as starting point even the experiments in random mutagenesis
might be more effective in order to achieve further improvements in production.

- 127 -
References
___________________________________________________________________________
6 References

Banik, A.K., Majumdar, S.K. (1974). Studies on Methionine Fermentation: Part I Selection of Mu-
tants of Micrococcus glutamicus & Optimum Conditions for Methionine Production. Indian Journal of
Experimental Biology vol.12 pp 363365.

Barrett G.C. (Editor) (1985) Chemistry and Biochemistry of the Amino Acids. Chapman and Hall
publishing house; first edition.

Battilotti, M., Barberini, U. (1988) Preparation of D-Valine from Isopropylhydantoin by Stereoselec-
tive Biocatalysis J.Mol.Cat. vol.43 pp. 343.

Becker, J., Klopprogge, C., Zelder, O., Heinzle, E., Wittmann, C. (2005). Amplified Expression of
Fructose 1.6-Bisphosphatase in Corynebacterium glutamicum Increases In Vivo Flux through the Pen-
tose Phosphate Pathway and Lysine Production on Different Carbon Sources. Applied and Environ-
mental Microbiology vol.71 (12) pp. 85878596.

Bendt, A.K., Burkovski, A., Schaffer, S., Bott, M., Farwick, M., Hermann, T. (2003). Towards a
phosphoproteome map of Corynebacterium glutamicum. Proteomics vol.3 pp 16371646.

Bonnarme, P., Amarita, F., Chambellon, E., Semon, E., Spinnler, H.E., Yvon, M. (2004). Methylthio-
acetaldehyde, a possible intermediate metabolite fort he production of volatile sulphur compounds
from L-methionine by Lactococcus lactis. FEMS Microbiology Letters vol.236 pp 85-90.

Blombach, B., Schreiner, M.E., Holatko, J., Bartek, T., Oldiges, M., Eikmanns, B.J. (2007). L-Valine
Production with Pyruvate Dehydrogenase Complex-Deficient Corynebacterium glutamicum. Appl.
and Environmental Microbiology vol.73 (7) pp.2079-2083.

Chmiel, H. (2006). Bioprozesstechnik. Published by Spektrum Akademischer Verlag., Second Edition.

Clement, Y., Laneelle, G. (1986). Glutamate Excretion Mechanism in Corynebacterium glutamicum:
Triggering by Biotin Starvation or by Surfactant Addition. Journal of General Microbiology vol.132
pp 925929.

Deutenberg, D. (2003) Genetische Optimierung der Methioninproduktion in Corynebacterium gluta-
micum ATCC 13032 Diploma thesis.

Eggeling, L., Sahm, H. (1999). L-Glutamate and l-lysine: traditional products with impetous develop-
ments. Appl. Microbiol. Biotechnol. vol.52 pp 146153.

Eggeling, L., Sahm, H. (2001). The cell wall barrier of Corynebacterium glutamicum and Amino Acid
Efflux. Journal of Bioscience and Bioengineering vol.92 (3) pp. 201-213.

Eggeling, L., Sahm, H. (2003). New ubiquitous translocators: amino acid export by Corynebacterium
glutamicum and Escherichia coli. Arch. Microbiol. vol.180 pp. 155 160.

- 128 -
References
___________________________________________________________________________
Furui, M., Yamashita, K. (1983). Pressurized reaction method for continuous production of L-alanine
by immobilized Pseudomonas dacunhae. J.Ferment.Technol. vol.61 pp 587-591.

Gerhardt, P., Murray, R.G.E., Costilow, R.N., Nester, E.W., Wood, W.A., Krieg, N.R., Phillips, G.B.
(1981) Manual of Methods for General Bacteriology. American Society for Microbiology

Gomes, J., Kumar, D. (2005). Methionine production by fermentation. Biotechnology Advances
vol.23 pp 4161.

Gomes, J., Kumar, D. (2005). Production of L-methionine by submerged fermentation: A review. En-
zyme and Microbial Technology vol.37 pp 3-18.

Haitani, Y., Awano, N., Yamazaki, M., Wada, M., Nakamori, S., Takagi, H. (2006). Functional analy-
sis of l-serine O-acetyltransferase from Corynebacterium glutamicum. FEMS Microbiol Lett. vol.255
pp 156163.

Hasegawa, H., Shinohara, Y., Akahane, K., Hashimoto, T. (2005). Direct Detection and Evaluation of
Conversion of D-Methionine into L-methionine in Rats by Stable Isotope Methodology. Journal of
Nutrition. vol.135 pp. 2001-2005.

Hashimoto, S., Katsumatra, R. (1994). L-alanine production by alanine racemase-deficient mutant of
Arthrobacter oxydans. Proc. Ann. Meet. Agric. Chem. Soc. (Japan). vol.341.

Heinrickson, R.L., Meredith, S.C. (1984). Amino Acid Analysis by Reverse-Phase High-Performance
Liquid Chromatography: Precolumn Derivatization with Phenylisothiocyanate. Analytical Biochemis-
try 136 pp 6573.

Hilliger, M. (1991). Biotechnologische Aminosureproduktion. Bio Tec vol.2 pp 40-44.

Husek, P., Simek, P. (2001). Advances in Amino Acid Analysis LCGC vol.19 (9).

Hwang, B.-J., Yeom, H.-J., Kim, Y., Lee, H.-S. (2001). Corynebacterium glutamicum utilizes both
transsulfuration and direct sulfhydrylation pathways for methionine biosynthesis. Journal of Bacteriol-
ogy vol.184 (5) pp 12771286.

Kase, H., Nakayama, K. (1974). Production of O-Acetyl-L-homoserine by Methionine Analog-
resistant Mutants and Regulation of Homoserine-O-transacetylase in Corynebacterium glutamicum.
Agr. Biol. Chem. vol.38 (10), pp 20212030.

Kase, H., Nakayama, K. (1975). O-Acetylhomoserine as an Intermediate in Methionine Biosynthesis
in Arthrobacter paraffineus, Corynebacterium glutamicum and Bacillus species. Agr. Biol. Chem.
vol.39 (3) pp 687-693.

Kase, H., Nakayama, K. (1975). L-Methionine Production by Methionine Analog-resistant Mutants of
Corynebacterium glutamicum. Agr. Biol. Chem. vol. 39 (1) pp 153160.

- 129 -
References
___________________________________________________________________________
Kase, H., Nakayama, K. (1975). Isolation and Characterization of S-Adenosylmethionine-requiring
Mutants and role of S-Adenosylmethionine in the Regulation of Methionine Biosynthesis in Coryne-
bacterium glutamicum. Agr. Biol. Chem. vol.39 (1) pp 161-168.

Keilhauer, C., Eggeling, L., Sahm, H. (1993). Isoleucine Synthesis in Corynebacterium glutamicum:
Molecular Analysis of the ilvB-ilvN-ilvC Operon. Journal of Bacteriology vol.175 (17) pp 55955603.

Keshavarz, K. (2003). Effects of Reducing Dietary Protein, Methionine, Choline Folic Acid and Vita-
min B
12
During the Late Stages of the Egg Production Cycle on Performance and Eggshell Quality.
Poultry Science vol.82 pp 14071414.

Kiefer, P., Heinzle, E., Zelder, O., Wittmann, C. (2004). Comparative Metabolic Flux Analysis of
Lysine-Producing Corynebacterium glutamicum Cultured on Glucose or Fructose. Applied and Envi-
ronmental Microbiology vol.70 (1) pp 229239.

Kinoshita, S., Ukada, S., Shimono, M. (1957). Studies on the amino acid fermentation. I. Production
of l-glutamic acid by various microorganisms J.Gen.Appl.Microbiol. vol. 3 pp 139205.

Kinoshita, S., Udaka, S., Akita, S. (1961). Method of producing l-glutamic acid by fermentation.
United States Patent Office, number 3.003.925 (Patented Oct. 10. 1961).

Kinoshita, S., Tanaka, K., Akita, S. (1961). Method of producing l-glutamic acid. United States Patent
Office, number 3.002.889 (Patented Oct. 3. 1961).

Kole, M.M., Thompson, B.G., Gerson, D.F., Senechai, J. (1986). Control of Nitrate Concentration in
Fermentations of Corynebacterium glutamicum. Biotechnology and Bioengineering vol.XXVIII pp
659662.

Koolman, J. and Rhm, K.-H. (1997). Taschenatlas der Biochemie. Stuttgart New York, Georg Thie-
me Verlag.

Kovaleva, G.Y., Gelfand, M.S. (2007). Regulation of Methionine / Cysteine Biosynthesis in Coryne-
bacterium glutamicum and related Organisms. Molecular Biology vol.41 pp 126 136.

Krmer, R. (1993). Mechanismen der Aminosuresekretion bei Corynebacterium glutamicum. Bioen-
gineering vol.9 pp 51-61.

Kumar, D., Garg, S., Bisaria, V.S., Sreekrishnan, T.R., Gomes, J. (2003). Production of methionine by
a multi-analogue resistant mutant of Corynebacterium lilium Process Biochemistry vol.38 pp 1165
1171.

Lee, H.-S., Hwang, B.-J. (2003) Methionine biosynthesis and its regulation in Corynebacterium glu-
tamicum: parallel pathways of transsulfuration and direct sulfhydrylation. Appl. Microbiol. Biotech-
nol. vol.62 pp 459467.

Lengeler, J.W., Schlegel, H.G., Drews, G. (1999). Biology of the Prokaryotes. Published by Blackwell
Science Inc., First Edition.

Leuchtenberger, W., Plcker, U. (1988). Herstellung von Aminosuren mit Hilfe biotechnologischer
Methoden. Chem.-Ing.-Tech. vol.60 (1) pp 1623.

- 130 -
References
___________________________________________________________________________
Leuchtenberger, W. (1996). Biotechnology, Volume 6, Products of primary metabolism: Amino Acids
Technical Production and Use VCH Weinheim publishing company.

Leuchtenberger, W., Huthmacher, R., Drauz, K. (2005). Biotechnological production of amino acids
and derivatives: current status and prospects, Appl. Microbiol. Biotechnol. vol.69 pp 1-8.

Liebl, W., Klamer, R., Schleifer, K.-H. (1989). Requirement of chelating compounds for the growth of
Corynebacterium glutamicum in synthetic media. Appl. Microbiol. Biotechnol. vol.32 pp 205-210.

Mampel, J., Schrder, H., Haefner, S., Sauer, U. (2005). Single-gene knockout of a novel regulatory
element confers ethionine resistance and elevates methionine production in Corynebacterium glu-
tamicum. Appl. Microbiol. Biotechnol. vol.68 pp 228236.

May, O., Nguyen, P.T., Arnold, F.H. (2000). Inverting enantioselectivity by directed evolution of hy-
dantoinase for improved production of L-methionine. Nature Biotechnology vol.18 (3) pp 317-320.

Misciatelli, L., Kristensen, V.F., Vestergaard, M., Weisbjerg, M.R., Sejrsen, K., Hvelplund, T. (2003).
Milk Production, Nutrient Utilization, and Endocrine Responses to Increased Postruminal Lysine and
Methionine Supply in Dairy Cows. J. Dairy Sci. vol.86 pp 275286.

Mondal, S. (1993). Influence of cystine on methionine production by Nocardia polychromogenes and
Brevibacterium ammoniagenes Research and Industry vol.38 p 101.

Mondal, S., Das, Y.B., Chatterjee, S.P. (1994). L-Methionine Production by Double Auxotrophic Mu-
tants of a Ethionine Resistant Straain of Brevibacterium heali. Acta Biotechnol. vol.14 (1) pp 61-66.

Mondal, S., Das, Y.B., Chatterjee, S.P. (1996). Methionine Production by Microorganisms. Folia Mi-
crobiol. 41 (6) pp 465 472.

Mondal, S., Das, Y., Samanta, T.K., Bhattacharya R., Chatterjee, S.P. (1990) L-Methionine production
by Nocardia polychromogenes and Brevibacterium ammoniagenes. Research and Industry vol.35 pp
11-14.

Mondal, S., Das, Y.B., Chatterjee, S.P. (1994) Improvement of L-methionine production by double
auxotrophic mutants of Brevibacterium heali LT
5
and LT
18
. Research and Industry vol.39 pp 239241.

Mondal, S., Chatterjee, S.P. (1994). Enhancement of Methionine Production by Methionine Analogue
Ethionine Resistant Mutants of Brevibacterium heali. Acta Biotechnol. vol.14 (2) pp 199-204.

Nakamura, J., Hirano, S., Ito, H., Wachi, M. (2007). Mutations of the Corynebacterium glutamicum
NCgl1221 Gene, Encoding a Mechanosensitive Channel Homolog, Induce L-Glutamic Acid Produc-
tion. Applied and Environmental Microbiology. vol.73 (14) pp 4491- 4498.

Nakayama, K., Araki, K. (1973). Process for producing L-Methionine. United States Patent Office
number 3.729.381. patented Apr.24.1973.

Naidanow, J., Minarikova, L., Barka, G., Fldi, P. (2005). Hochauflsende Aminosureanalyse durch
Kapillar-HPLC. LaborPraxis 6 (5) pp 56-59.

Noftsger, S., St-Pierre, N.R. (2003). Supplementation of Methionine and selection of Highly Digesti-
ble Rumen Undegradable Protein to Improve Nitrogen Efficiency for Milk Production. J. Dairy Sci.
vol.86 pp 958969.

Pack, M. (2004). Aminosuren in der Tierernhrung. Degussa Science Newsletter vol.06.

- 131 -
References
___________________________________________________________________________
Petritis, K., Elfakir, C., Dreux, M. (2002). A comparative study of commercial liquid chromatographic
detectors for the analysis of unmodified amino acids. Journal of Chromatography A vol.961 pp 921.

Rckert, C., Phler, A., Kalinowski, J. (2003). Genome wide analysis of the l-methionine biosynthetic
pathway in Corynebacterium glutamicum by targeted gene deletion and homologous complementa-
tion. Journal of Biotechnology vol.104 pp 213228.

Schlenk, F., Hannum, C.H., Ferro, A.J. (1978). Biosynthesis of Adenosyl-D-methionine and Adeno-
syl-2-methylmethionine by Candida utilis. Archives of Biochemistry and Biophysics vol.187 (1) pp
191196.

Schmid, R.D. (2002). Taschenatlas der Biotechnologie und Gentechnik. Wiley VCH publishing
company., First Edition.

Sharma, S. and Gomes J. (2001). Effect of Dissolved Oxygen on Continous Production of Methionine
Eng. Life Science 1 (2) pp 6971.

Sharma, S. (2001). Strain improvement and reactor studies for the production of l-methionine by
Corynebacterium lilium Dissertation Indian Institute of Technology, Delhi.

Shiio, I., Ujigawa, K. (1980). Presence and regulation of -ketoglutarate dehydrogenase complex in a
glutamate producing bacterium, Brevibacterium flavum. Agric.Biol.Chem. vol.42 pp 1897-1904.

Stahl, E. (1967). Dnnschicht-Chromatographie ein Laboratoriumshandbuch. Springer Verlag Ber-
lin, 2
nd
edition.

Stansen, K.C. (2005). Charakterisierung der Ausscheidung von L-Glutamat bei Corynebacterium glu-
tamicum. Dissertation Forschungszentrum Jlich 2005.

Stehr, R. (1996). Optimierung der Herstellung von L-Methionin aus D,L-Methylthioethylhydantoin
mit Arthrobacter sp. DSM 7330. Dissertation University of Braunschweig.

Tani, Y., Lim, W.-J., Yang, H.-C. (1988). Isolation of L-Methionine-Enriched Mutant of a Methy-
lotrophic Yeast, Candida boidinii No.2201. J.Ferment.Technol. vol.66 (2) pp 153-158.

Tokuyama, S., Hatano, K. (1996). Overexpression of the gene for N-acylamino acid racemase from
Amycolatopsis sp. TS-1-60 in Escherichia coli and continous production of optically active methion-
ine by a bioreactor. Appl. Microbiol. Biotechnol. vol.44 pp 774777.

Toride Y. (2002). Lysine and other amnio acids for feed: production and contribution to protein in
animal feeding. FAO Expert consultation and Workshop for Protein Sources for the Animal Feed In-
dustry, WHO & International Feed Industry Federation, Bangkok, Thailand.

Tosaka, O., Takinami, K. (1986). In: Aida, K., Chibata, I., Nakayama, I., Takinami, K., Yamada, H.,
editors. Biotechnology of amino acid production 1986: pp 152172.

Trtschel, C. (2005) Methioninaufnahme und Export in Corynebacterium glutamicum Dissertation
University of Cologne.

Trtschel, C., Deutenberg, D., Bathe, B., Burkovski, A., Krmer, R. (2005) Characterization of Me-
thionine Export in Corynebacterium glutamicum. Journal of Bacteriology vol.187 pp 37863794.

- 132 -
References
___________________________________________________________________________
Voelkel, D. (1993). Produktion von L-Methionin mit neuen Mikroorganismen nach dem Hydantoina-
se-Verfahren und Aufklrung des enzymatischen Mechanismus. Dissertation University of Braun-
schweig.

Voet, D.; Voet, J.G., Pratt, C.W. (2002). Biochemie. Published by Wiley - VCH Weinheim. Second
Edition.

Wartenberg, A. (1989) Einfhrung in die Biotechnologie. Gustav Fischer Verlag, Stuttgart.

Wittmann, C., Kiefer, P., Zelder, O. (2004). Metabolic Fluxes in Corynebacterium glutamicum during
Lysine Production with Sucrose as Carbon Source. Applied and Environmental Microbiology vol.70
pp 7277-7287.

Wbbeking, C. (2005). Entwicklung analytischer Methoden zum effizienten Screening methioninpro-
duzierender Mikroorganismen. Diploma thesis Bundesforschungsanstalt fr Landwirtschaft (FAL),
Braunschweig.

Yahata, S., Tsutsui, H., Yamada, K., Konehara, T. (1993). Fermentative production of D-alanine.
Proc. Ann. Meet. Agric. Chem. Soc. (Japan) vol.92.

- 133 -
Abbreviations
___________________________________________________________________________
7 Abbreviations

ACQ 6-amino-quinolyl-N-hydroxy-succinimidyl-carbamate
ATCC American Type Culture Collection
BDM Bio dry mass
BY Bouillon yeast extract
BSA Bovine serum albumin
CER Carbon dioxide evolution rate
DSM German Collection of Microorganisms
EMS Ethyl-methanesulfonate
GC Gas chromatography
HPLC High performance / pressure liquid chromatography
IMH Indolyl-methyl-hydantoin
kDa Kilodalton
MOPS Morpholine propane sulphonic acid
MS Mass spectometry
MTEH Methyl-thio-ethyl-hydantoin
nm Nanometer (wavelength)
n Stirring speed
NTG N-methyl-N-nitro-nitrosoguanidine
OD Optical density
OPA Ortho-phtaldialdehyde
OUR Oxygen uptake rate
pO
2
Oxygen partial pressure
PITC Phenylisothiocyanate
q
CO2
= CER (based on molar mass)
q
O2
= OUR (based on molar mass)
RQ Respiratory coefficient
THF Tetra hydro folate
TLC Thin layer chromatography
TOF Time of Flight
Y
X/S
Yield coefficient [(g) bio dry mass / (g) substrate]
Y
P/S
Yield coefficient [(g) product / (g) substrate]
Y
P/X
Yield coefficient [(g) product / (g) bio dry mass]
UV Ultraviolet light

- 134 -
Attachment
___________________________________________________________________________
8 Attachment

Table A.1: Additional amino acids in minor concentrations measured in the cultivation with KH1 +
KH2 medium (see Figures 4.09 + 4.10; chapter 4.4.1).

KY10574
Time (h)
L-leucine
(mg/l)
L-iso-leucine
(mg/l)
L-valine
(mg/l)
L-glutamic acid
(mg/l)
0 42.3 17.7 21.1 6.3
24 46.7 20 27.04 15.8
48 40.8 16.8 30.1 28.8
72 16.8 5.1 40 29.4
96 19.6 7.9 37.8 35.3
120 66.1 30.7 68.03 64.5
144 74.29 39.6 76.2 65.9
168 69.3 36.6 79.3 65.27
240 9.3 5.2 28.4 18.1

Figure A.1: Diagram of the GC measurement (performed at the Institute of Organic Chemistry, TU
Braunschweig) and the amino acid identification of the sample after 96 h cultivation time from the
cultivation with KH2 medium showed in the results in the Figure 4.10 (chapter 4.4.1).

- 135 -
Attachment
___________________________________________________________________________

Figure A.2: Diagram of the GC measurement (performed at the Institute of Organic Chemistry, TU
Braunschweig) of the reagent 3B from the derivatization kit for amino acids (Phenomenex company)
to identify the important substances for the modification of the amino acids, in this case the catalysator
3-methyl pyridine (see chemical reaction in chapter 4.4.1.1 and Table 3.10).

Figure A.2: Diagram of the GC measurement (performed at the Organic Chemistry, TU Braun-
schweig) of the reagent 4B from the derivatization kit for free physiological amino acids (Phenomenex
company) to identify the important substances for the modification of the amino acids, in this case the
propyl chloro formate substances (see chemical reaction in chapter 4.4.1.1 and Table 3.10).

- 136 -
Attachment
___________________________________________________________________________

L-glutamic acid
(modified)
Figure A.3: Diagram of the mass spectrometry fragments (MS; performed at the Organic Chemistry,
TU Braunschweig) of modified L-glutamic acid (see Figure 4.15) obtained through GC-MS from the
sample taken after 120 h cultivation time (for the cultivation see Figure 4.10 / chapter 4.4.1).

L-lysine
(modified)
TU Braunschweig) of modified L-lysine (see Figure 4.15) obtained through GC-MS from the sample
taken after 120 h cultivation time (for the cultivation see Figure 4.10 / chapter 4.4.1).

- 137 -
Attachment
___________________________________________________________________________

glycine
(modified)
TU Braunschweig) of modified glycine, modified L-leucine and modified L-iso-leucine (see Figure
4.15) obtained through GC-MS from the sample taken after 120 h cultivation time (for the cultivation
see Figure 4.10 / chapter 4.4.1).

L-alanine
(modified)
L-iso-leucine
(modified)
L-leucine
(modified)
TU Braunschweig) of modified L-alanine (see Figure 4.15) obtained through GC-MS from the sam-
ple taken after 120 h cultivation time (for the cultivation see Figure 4.10 / chapter 4.4.1).

- 138 -
Attachment
___________________________________________________________________________

L-valine
(modified)

TU Braunschweig) of modified L-valine (see Figure 4.15) obtained through GC-MS from the sample
taken after 120 h cultivation time (for the cultivation see Figure 4.10 / chapter 4.4.1)

- 139 -
Curriculum vitae
___________________________________________________________________________
Curriculum vitae / Lebenslauf

Persnliche Daten

Name: Hajo Kampe Reershemius
Geburtsdatum / -ort: 30.08.1978 in Norden (Ostfriesland)
Staatsangehrigkeit: Deutsch
Adresse: Nordstrasse 29, 38106 Braunschweig
Familienstand: ledig

Promotion

02.2005 06.2008 Wissenschaftlicher Mitarbeiter zum Zwecke der Promotion am
Institut fr Biochemie und Biotechnologie, TU Braunschweig

Studium

10.1999 01.2005 Studium der Biotechnologie (Diplom)
Technische Universitt Braunschweig

27.09.2001 Vordiplom Biotechnologie

11.01.2005 Diplom Biotechnologie (Dipl.-Biotechnol.)

Schulausbildung

08.1985 07.1989 Grundschule Krummhrn Sd

08.1989 06.1991 Orientierungsstufe Emden Wybelsum

08.1991 06.1998 Johannes Althusius Gymnasium Emden

22.06.1998 Abitur

Ttigkeiten

08.1998 08.1999 Zivildienst beim Deutschen Roten Kreuz
Kreisverband Emden

- 140 -

L Methionine PDF

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

L Methionine PDF

Hochgeladen von

Copyright:

Verfügbare Formate

Production of L-methionine with

Das könnte Ihnen auch gefallen